Topping E30 DAC Review

[modus2000]

should I worry about the jitter if I'm going to use ONLY toslink ?

I'm looking to 2 models E30II or SMSL SU-1
Would be SU-1 better ? Or it has another issues?

 

[antcollinet]

should I worry about the jitter if I'm going to use ONLY toslink ?

I'm looking to 2 models E30II or SMSL SU-1
Would be SU-1 better ? Or it has another issues?

No - the jitter is one of those things that is measurably bad compared to some other DAC's - but is still not bad enough to be audible.

 

[PyramidElectric]

RE Airport Express, not sure which model you have but if it's the latest one then then the toslink output on it is indeed quite jittery/unstable, I have a couple of SMSL dacs (DO100 and D-6s) that have a DPLL setting available to accomodate sources such as this (and 'bad' TVs etc) and they both need setting to '9' (the maximum/most accomodating) to avoid dropouts. Having said that, at this setting all is good. So yeah, the Jitter of the DAC is probably nothing to worry about, but the Jitter of the source, and whether the Toslink reciever can lock onto it might be the thing to worry about.

 

[jmb239]

I haven't had any jitter issues with Airport Express and E30. I'm running the last model Airport Express with 7.6.9 firmware into an E30. This makes it Airplay 1.

The latest firmware, 7.8.1, is Airplay 2. 7.8.1 adds distortion that is audible when listening on headphones. This is a known issue, but I don't think it is jitter related.

 

[modus2000]

I haven't had any jitter issues with Airport Express and E30. I'm running the last model Airport Express with 7.6.9 firmware into an E30. This makes it Airplay 1.

I did use latest version of firmware (A1392) and it was lossy! Thank you for this trick!!!

My brain say "buy SMSL because it has better Jitter measurement"
but my heart say "buy Topping E30II Lite - although I'm going to use Toslik only, I already have L30 amp which I can pair sometime to E30II"

 

[westonlast]

I know I should provide blind test results. Maybe I'll do some rigorous testing at some point.

I've read all the stuff about PCM vs. DSD. I agree that, in theory, neither is better than the other, from an audibility perspective. In practice, perhaps it depends on implementation.

That's why I'm posting this here and not on a general "let's talk about DSD" thread. I think what I'm hearing is due to this particular DAC's implementation of PCM vs. DSD.

I've had this DAC for a while, just listening to PCM, and I was happy with it. But curiosity started to kill the cat. I do all my listening via foobar2000. I've decided to dabble in https://sourceforge.net/projects/sacddecoder/files/foo_dsd_processor/ .

The past couple days I've configured foobar2000 to convert 44100/16 and 96000/24 to DSD128. Been flipping back and forth to PCM, of course; who wouldn't. Not particularly quickly or rigorously, though. I'm fairly certain DSD128 sounds better. Unlike the countless others who have said the same of DSD, I'll actually describe the difference, and even in non-romantic language:

I can hear around the edges of the sounds more, but it doesn't sound harsh or edgy. It sounds more natural. I think the PCM implementation is blurring the sound a bit.

Like loose shrink-wrap has been shrunk, so all you see is the object and not the shrink-wrap.

Now that subjective BS is done, back to the technical talk. Looking through the datasheet of the AKM chip, it says the DSD processing respects the SACD specification by putting full-scale at 25% - 75% of the duty cycle. Or something, just my vague understanding. Maybe that's a reason DSD is sounding better; it's more quiet than full-scale PCM. Maybe that means the DSD has more room to "breathe" in terms of the output voltages and the reconstruction filter. Although, I am applying attenuating EQ as a DSP (before the DSD conversion, obviously), so that should make any kind of clipping issues even more unlikely.

I'm using the E30 with the HEDD Audio HEDDphone, which should help me hear a difference, if there is any. Also, I'm using the USB data input and power from the computer via the included USB-to-barrel connector as well.

A flaw of the E30 is the popping on any manual seek or track change. Even when starting a DSD track from nothing playing, the damn thing pops/cracks. DSD is required to be output to the ASIO driver (Windows' outputs don't support DSD, unless DSD-over-PCM [DoP] is used, which I haven't tried and adds extra complication), as opposed to the normal Windows DirectSound output, so each manual operation, in foobar2000 anyway, always causes the device to re-init/re-sync.

I imagine it should be possible for the foobar2000 DSD components to monitor the DSD "waveform" to know when digital silence is happening and wait to change until that. And also, foobar2000 could allow seeks and track changes to not re-init the device. With these two fixes, the E30's popping, which is documented in the AKM datasheet as a fact of life, by the way, could be avoided.

I'm listening to PCM using the default filter 3, and I'm listening to DSD using the default filter 1. The DSD default stopband starts at 30 kHz.

I've had this device for years, by the way. Now I'm enjoying it even more, since I think it sounds even better.

 

[antcollinet]

I know I should provide blind test results.

The fact that you haven't completely invalidates the rest of the post.

Your sighted listening impressions are, frankly, pointless.

And you can't speculate, guess, or intuit engineering.

 

[Sokel]

I know I should provide blind test results. Maybe I'll do some rigorous testing at some point.

I've read all the stuff about PCM vs. DSD. I agree that, in theory, neither is better than the other, from an audibility perspective. In practice, perhaps it depends on implementation.

That's why I'm posting this here and not on a general "let's talk about DSD" thread. I think what I'm hearing is due to this particular DAC's implementation of PCM vs. DSD.

I've had this DAC for a while, just listening to PCM, and I was happy with it. But curiosity started to kill the cat. I do all my listening via foobar2000. I've decided to dabble in https://sourceforge.net/projects/sacddecoder/files/foo_dsd_processor/ .

The past couple days I've configured foobar2000 to convert 44100/16 and 96000/24 to DSD128. Been flipping back and forth to PCM, of course; who wouldn't. Not particularly quickly or rigorously, though. I'm fairly certain DSD128 sounds better. Unlike the countless others who have said the same of DSD, I'll actually describe the difference, and even in non-romantic language:

I can hear around the edges of the sounds more, but it doesn't sound harsh or edgy. It sounds more natural. I think the PCM implementation is blurring the sound a bit.

Like loose shrink-wrap has been shrunk, so all you see is the object and not the shrink-wrap.

Now that subjective BS is done, back to the technical talk. Looking through the datasheet of the AKM chip, it says the DSD processing respects the SACD specification by putting full-scale at 25% - 75% of the duty cycle. Or something, just my vague understanding. Maybe that's a reason DSD is sounding better; it's more quiet than full-scale PCM. Maybe that means the DSD has more room to "breathe" in terms of the output voltages and the reconstruction filter. Although, I am applying attenuating EQ as a DSP (before the DSD conversion, obviously), so that should make any kind of clipping issues even more unlikely.

I'm using the E30 with the HEDD Audio HEDDphone, which should help me hear a difference, if there is any. Also, I'm using the USB data input and power from the computer via the included USB-to-barrel connector as well.

A flaw of the E30 is the popping on any manual seek or track change. Even when starting a DSD track from nothing playing, the damn thing pops/cracks. DSD is required to be output to the ASIO driver (Windows' outputs don't support DSD, unless DSD-over-PCM [DoP] is used, which I haven't tried and adds extra complication), as opposed to the normal Windows DirectSound output, so each manual operation, in foobar2000 anyway, always causes the device to re-init/re-sync.

I imagine it should be possible for the foobar2000 DSD components to monitor the DSD "waveform" to know when digital silence is happening and wait to change until that. And also, foobar2000 could allow seeks and track changes to not re-init the device. With these two fixes, the E30's popping, which is documented in the AKM datasheet as a fact of life, by the way, could be avoided.

I'm listening to PCM using the default filter 3, and I'm listening to DSD using the default filter 1. The DSD default stopband starts at 30 kHz.

I've had this device for years, by the way. Now I'm enjoying it even more, since I think it sounds even better.

I would give your post some merit if your AKM implementation was suffering from an ESS hump for example, such as this one:

PCM vs DSD

Did a little comparison and I was wondering if any friend with better gear here has tested this. Results you're about to see is probably narrowed to the specific DAC but it's interesting to see how it handles things differently. I will not try to decipher the results other that DSD seems much...

www.audiosciencereview.com

The above gets a ~30dB IMD improvement using DSD, right at the (nicely made) recording usual levels, so it could be considered a (measurable, and that's all that matters from an engineering point of view) major improvement.

But AKM, I don't think so, their long tradition (at least the one I value them for) is suggesting no BS chips who measure decent across ALL spectrum, levels (VERY important) and conditions in general.

 

[westonlast]

The fact that you haven't completely invalidates the rest of the post.

No shit there's no objective evidence. But it doesn't invalidate my post. Meanie. If you actually read it.

And you can't speculate, guess, or intuit engineering.

I can do whatever I want.

Stupid me thinking Internet people could read a post written in English objectively.

 

[Sal1950]

I can do whatever I want.

You can, but that doesn't make you correct.

 

[westonlast]

You can, but that doesn't make you correct.

Never said I was. Work on reading comprehension.

 

[antcollinet]

No shit there's no objective evidence. But it doesn't invalidate my post. Meanie. If you actually read it.


I can do whatever I want.

Stupid me thinking Internet people could read a post written in English objectively.

It does invalidate the conclusions you draw in your post based on unsighted listening.


And when I said 'you cant' I didn't mean you are not allowed - of course from that point of view you can do what the hell you like. My meaning was "It is not possible"


You either know and understand - or you don't. Guessing doesn't work. It is as often as not, counter intuitive.