[BrotherKathos]

 

[BrotherKathos]

That makes sense then. AKM dacs are known for their 'velvet sound' which smooths out highs and is not the most objective of dacs. Multibit dacs are also not as revealing as delta sigma. That's why your magni 3 works well with them. Get a highly objective dac like any sabre 9038 based chip and you will meet the real magni 3.

 

[BrotherKathos]

Was curious about the Massdrop THX 789 amp and was comparing output power with the Magni 3 (using mfg provided specs on Massdrop.com and Schiit.com), and was surprised to see Magni 3 specs more power into single ended:

@16ohms they are tied at 3 watts
@32 ohms Magni 3 is 2 watts, THX 789 1.8 watts
@300 ohms Magni 3 is 430mW, THX 789 200mW
@600 ohms Magni 3 is 230mW, THX 789 100mW

According to these specs the Magni 3 power advantage increases as drive conditions become more difficult...a good thing. Of course each mfg may have used different standards to establish clipping levels, and only results using the same protocol by an unbiased tech would be definitive. Even so all the hoo hah about the THX 789 being so powerful seems overstated, particularly with the Magni 3 on the scene.

And the differences in distortion specs between the two units should be well below thresholds for audibility, leading me to think under well controlled ABX conditions, using single ended output on the 789, there will be no aural difference between the two units.

You are only looking at SE performance. The 789 is more powerful than the magni 3 in balanced, which is what the vast majority of the people buying this amp will most likely use. Its what I'll be using with my Audeze LCD-2C.

 

[tamleo]

I don't think my Mymby are less revealing than any other ds dacs (including the ds Gungnir).But I would love my Mymby's bass to be a little less bloated.
I have been listening to my "unicorn" LCD2 with my M3 recent days, imo the M3 lacked of some high treble and sometimes i felt the pairing too dark. So i came back using My LCD2 with my M2u and liked it better. The m3 is for my balanced amateur iem now.

 

[treecloud]

You are only looking at SE performance. The 789 is more powerful than the magni 3 in balanced, which is what the vast majority of the people buying this amp will most likely use. Its what I'll be using with my Audeze LCD-2C.

Yes of course I'm looking at single ended performance, to compare one single block of amplification to another. Otherwise it's not an "apples to apples" comparison. If Schiit made a balanced version of the Magni 3, again, going by mfg specs which may have used different test protocols, it seems that it would be more powerful than the 789.

In most cases, when well designed amps are kept within respective power envelope, objective listening tests do not show aural benefit to balanced amplification. I'd be curious to know what percentage of Head-Fiers used balanced cans.

 

[Tuneslover]

Yes of course I'm looking at single ended performance, to compare one single block of amplification to another. Otherwise it's not an "apples to apples" comparison. If Schiit made a balanced version of the Magni 3, again, going by mfg specs which may have used different test protocols, it seems that it would be more powerful than the 789.

In most cases, when well designed amps are kept within respective power envelope, objective listening tests do not show aural benefit to balanced amplification. I'd be curious to know what percentage of Head-Fiers used balanced cans.

I do, HE500 and HD650 with Mimby & Jotunheim

 

[riffrafff]

The m3 is for my balanced amateur iem now.

I used to be a balanced amateur; now I'm a balanced professional.

 

[treecloud]

I do, HE500 and HD650 with Mimby & Jotunheim

The HeadFi community would probably have the highest % of balanced can users on the planet. Perhaps there's a rough correspondence between that ratio to the ratio of balanced to single ended only headphone amps. Guesses? 20%?

 

[NA Blur]

I still don't think it's biased towards sounding brighter. Perhaps better and more linear extension in the treble, but I challenge you to find me the measurements that hold up the treble boost claim.

Where do you see that it was "engineered that way?"

 

[BrotherKathos]

I still don't think it's biased towards sounding brighter. Perhaps better and more linear extension in the treble, but I challenge you to find me the measurements that hold up the treble boost claim.

Where do you see that it was "engineered that way?"

https://www.audiosciencereview.com/...g-a30-and-schiit-magni-3-headphone-amps.4024/

Here is some measurements. The reviewer also found the magni 3 to be bright. He attributed it to the magni 3's drastic rise in distortion compared to the topping a30 a you raise the volume, possibly adding high frequency artifacts that make the magni 3 brighter. I don't know about all that, but the magni 3 is absolutely brighter on every piece of audio equipment I have period. Maybe your equipment works great with the magni 3, but I buy mostly objective oriented dacs and amps so the magni 3 is just not my cup of Joe. Its good enough for me with my iphone at least, so I'll use it until I get the THX AAA and then either move it to another house or dump it on the used market.

 

[BrotherKathos]

I still don't think it's biased towards sounding brighter. Perhaps better and more linear extension in the treble, but I challenge you to find me the measurements that hold up the treble boost claim.

Where do you see that it was "engineered that way?"

I also wanted to note that the dac and headphones he used in the magni 3 review are basically identical to my setup. He was using a topping DX7 and the Hifiman He-400i. I'm also using the 400i and a Topping D50 which uses the same dac chips and is basically the same unit without a head amp or balanced outs. The D50 actually measures better than the Dx7. I wish I would have seen this review before I bought the magni 3. It would have saved me the $115, but oh well. Like i said before, its fine for a bedside setup with the iphone ntil the O2 amp replaces it when the THX AAA amp gets here in February.

 

[NA Blur]

Now I can see a DAC sounding brighter. There is all sorts of filtering going on especially if the DAC is DSD capable for higher bit depths.

I read those measurements too and I don't see a single FR measurement anywhere. Did you read them?

 

[BrotherKathos]

Now I can see a DAC sounding brighter. There is all sorts of filtering going on especially if the DAC is DSD capable for higher bit depths.

I read those measurements too and I don't see a single FR measurement anywhere. Did you read them?

This was a comparison. The magni 3 was brighter than the topping a30 and had dirtier power. Whether or not that makes it brighter than other amps, or whether its as you say and its the dac, then who knows. If my dacs are bright, then my O2 amp and JDS Labs EL Amp and Audioengine D1 are very dark, and my iphone is so dark as to have no high frequency at all. I'm just stating my experience, that also matches with others using similar equipment that is not present on known neutral gear. I've even heard people say the O2 amp is bright, and if that is the case then the magni 3 is shrill by comparison. My O2 amp and EL amp are way more neutral than my magni 3, just the facts. I don't need a measurement to tell me what I'm hearing. Maybe I just got a schiit unit>>>

 

[joeexp]

https://www.audiosciencereview.com/...g-a30-and-schiit-magni-3-headphone-amps.4024/

Here is some measurements. The reviewer also found the magni 3 to be bright. He attributed it to the magni 3's drastic rise in distortion compared to the topping a30 a you raise the volume, possibly adding high frequency artifacts that make the magni 3 brighter. I don't know about all that, but the magni 3 is absolutely brighter on every piece of audio equipment I have period. Maybe your equipment works great with the magni 3, but I buy mostly objective oriented dacs and amps so the magni 3 is just not my cup of Joe. Its good enough for me with my iphone at least, so I'll use it until I get the THX AAA and then either move it to another house or dump it on the used market.

There is this guy called "amirm" who has got a personal and well documented problem with Schiit products.
I wouldn't take his "reviews" seriously.

 

[treecloud]

There is this guy called "amirm" who has got a personal and well documented problem with Schiit products.
I wouldn't take his "reviews" seriously.

He brings some balance to the audio world by introducing some objectivity, but he is not really all that objective either really. He appears not to know much about psycho-acoustic research on threshold for audibility of the common distortions we measure of electronic audio gear. Generally speaking, thresholds are much higher than the levels he is typically comparing, but he does not go to great lengths to point out these differences are thought to be inaudible. Tests for distortion are not completely meaningless, it's interesting to know where the theoretical limit is currently. But in terms of sound quality they may be completely meaningless.

In addition, if he wanted to be truly objective he would mount complex time consuming aural tests with a variety of listeners including professional audiophiles, using ABX protocols. Pretty sure he's not interested in adding that layer of complexity to what he's already doing. Meanwhile, to his credit, he does give lip service to the problem from time to time saying something to the effect "these distortion differences are probably inaudible".

And then there's the big elephant in the room we can't measure and therefor no one is talking about, and that are the distortions that remain in the best systems that tell us we are listening to reproductions of music, not live music. We are getting closer, true, but still miles from true convergence. What are those unmeasured distortions? Mostly compression of "reality" that take various forms. "Waveform launch" is a biggie. Scale distortions. "Shape" (of aural phenomena) distortions. Dynamic and frequency range compressions. Dimensionality distortions.

These distortions in aggregate may be as high as (let's say) 70%. Not 0.00001%. Some "objectivists" give objectivity a bad name