Citypulse DA2.03e

[Spare Tire]

Hi everyone, i just recieved the DA 2.03e. The manual is puzzling me with regards to the upsampling buttons. Here's what the manual says:
Fs-AUTO: When the "Fs-AUTO" is on, it is the condiction of random frequency.[sic]
88.2KHz: When the "88.2KHz" is on, it is the condiction of random frequency: the signal frequency is 44.1Khz the output frequency is 88.2KHz.[sic]
96KHz: When the "96KHZ" is on, it is the condiction of random double frequency: the signal frequency is 48Khz the output frequency is 96KHz.[sic]

That's what's written word for word. Now, what that COULD mean is:
1. both 88.2 and 96.0 will upsample any frenquency by x2 (they do the same thing)
2. both 88.2 and 96.0 will upsample only 44.1KHz and 48KHz to 88.2 and 96.0 respectively
3. 88.2 upsamples anything to 88.2 only and 96.0 will upsample anything to 96.0 only
4. Fs-AUTO resamples everything to 44.1KHz
5. Fs-AUTO resamples everything to 48KHz
6. Fs-AUTO does not resample
Moreover, the dac is said to support up to 192KHz signal, but i can't very that since i don't have anything that outputs at more than 96 right now.

I've tried to verify some of these with hypothesis by using the upsampling function of foobar. However the DAC uses native USB Audio driver from windows and foobar cannot make it go through Kernel Streaming, so i cannot guarantee the signal will go through at that sample rate.
Here's what i've heard using foobar resampler to try replaying a 44.1KHz file (udial) at various sample rates, i did this with the front phone out to my grado SR-60, since the DAC didn't come with any RCAs and i don't have any handy to plug into my staxes.

UPDATE 4: In light of recent findings that the USB can only output 32-48kHz 16bit, all the observations i made that weren't in that range were being resampled by windows before being fed to the DAC. That explains why everything above 48 sounded the same as 48. I've set up ASIO4ALL to play with foobar and these are the results. Everything that was not in the range of 32-48kHz prompt a popup error and would not play.

with the DAC set to Fs-AUTO:
32000: normal
44100: normal
48000: normal

with the DAC set to 88.2KHz:
32000: some static in the left channel, sounds like a geiger counter
44100: whacky electronic stuff in the last 3 tones, like a UFO (lets call it w1)
48000: whacky electronic stuff in the last 3 tones, like a police syren thing (w2), surprisingly no geiger counter noise
Asside from udial, but udial included, all the music have a very slight aura of metalic ring around them. It's more evident with udial, a lot less evident with complexe music.

with the DAC set to 96.0KHz:
32000: normal like Fs-AUTO
44100: geiger counter static in the left channel only, sounds radioactive, also some pops in both channels, whacky electronic police syren but different from last one (call it w3)
48000: no geiger static, just isolated pops in both channels, whacky electronic phaser blast (call it w4)
Aside from udial, when i play music with the DAC set to 96.0KHz and foobar not resampling (effectively 44.1KHz) i get the radioactive geiger counter pops in the left channel. It is VERY evident and totally unbearable. Additionnally, there is the same metalic ring as with 88.2, a little more of it.

From all the hypothesis i have above, i can conclude some of these things:
1. Not true since 88.2 and 96 sound different.
2. Not true since 88.2 and 96 on the DAC sounded different with different foobar resamplings, especially at the 3 middles sample rates of 32kHz, 44.1kHz and 48kHz.
3. Not true, if 88.2 on the DAC only resampled to anything to 88.2 then Fs-AUTO with foobar at 88200 should sound the same as the DAC at 88.2 and foobar not resampling but it is not the case, same goes for 96 (this conclusion supposes that the DAC upsampling 44.1 signal to 88.2 and foobar upsampling 44.1 signal to 88.2 sound the same, which isn't necessarily the case, so you could debate this conclusion).
If the we suppose that foobar's upsampling and the DAC's upsampling might sound slightly different then this is the only plausible explanation, considering hypothesis 1 and 2 are (sonically) without doubt false.
So hypothesis 3 is a maybe.

4. Probably false, since withing Fs-AUTO, different foobar resampling gives different sound. If Fs-AUTO resampled everything to 44.1kHz, then all foobar resampling should sound the same.
The different sound might be imputable to the foobar resampler (and it certainly might) so perhaps i should try a file with a different native sample rate. But then again, the equipment used to record that some, or the mastering process, or whatever, would have made different sample rates sound different too. Perhaps try it out with two files of native sample rates that are very very high, where people say it becomes overkill and the differences become irrelevant. Anyways, this one is inconclusive.
5. Inconclusive for the same reason as 4.
6. Inconclusive for the same reason as 4, but i think this one is probably true. Maybe you could tell without a doubt if you looked at the actual circuit, but i know nothing of such.

There you have it, i'm still as confused as ever about the upsampling. Anybody care to help me figure out what the heck is going on, please do. Another thing, about the "geiger counter radioactive pops", there isn't any audible with the 88.2KHz upsample, even if it's slightly more harsh and sharp and the soundstage sound a bit deeper because the echoes are sharper. But with the dac set to 96.0KHz, we get this radioactive pops in the left channel with any imput at 44.1KHz. I don't know why, it really pisses me off. I'll contact the seller (obad_imports) see if he can reproduce the problem or if i just have a defective unit. Since i guess i'm the first on Head-Fi with this unit, nobody can help me, and i doubt people will buy if they hear this problem from me. I'd ask eddie too but i didn't buy the unit from him so it'd be abusing....
I still haven't tried them out with my Staxes, which i listen most of my music with, so i can't really tell about the sound quality, but so far, it sounds nice enough. Haven't done extensive comparative listening against the Sonica yet so i can't tell about that either, just off the top of my head i think it is a bit better (sure hope so! given the price difference).

UPDATE: I installed ASIO4ALL and saw that the windows usb audio driver only supports sample rates from 32-48kHz 16bit. Anything outside of that range dies. Major bummer.
UPDATE 2: I tried to load my sonica with microsoft's USB audio driver and in ASIO4ALL, it says "USB Audio Device" and gives the correct frequency range support for the sonica, that is from 8-96kHz 24bit. The citypulse displays "USB Audio CODEC" instead of "USB Audio Device". Does anyone know what this means? Please help.
UPDATE 3: Kernel Streaming in foobar gives a popup error at 24 and 32bits, 16bit doesn't popup any error but doesn't play.
UPDATE 5: The official stance (from eddie) is that upsample to 88.2 from 44.1 only, and to 96 from 48 only.

[ezkcdude]

Sorry, but that post was indecipherable. Are you translating from chinese or something? Is "condiction" a word I've never heard of?

[Spare Tire]

Yep, it's copied word for word, and with all their spelling mistakes.

[Spare Tire]

I am beginning to think the USB input really doesn't take anything else than 32-48kHz 16bit signal. With ASIO or kernel streaming, anything outside of that range wont pass. That explains some of the observations on my listening test of udial playing through foobar resampler. Actually, only the 32-48kHz signal passed through and all the others were resampled by windows for the DAC to accept. That explains why anything resampled above 48 sounds the same as 48, even 88.2 that i expected to sound like 44.1 instead of 48. I'll make the experience once more with ASIO later.

Asside from that, the unit sounds really good late at night, which probably means it's sensitive to interference and my appartment must have plenty of it. Also it's a big bummer that the USB will only accept 32-48kHz 16bit. Actually i don't have any music that isn't in that range, but this is not very future proof. DVD-A wont be able to play through the usb input. Why oh why did they botch the USB in, might as well have bought a DA7.2x for 100$ less.

[laxx]

Kernel Streaming doesn't work for USB, at least it doesn't for myself and a bunch of other people here.

It's limited to 48khz, unless otherwise stated (Benchmark Dac1) or it uses custom drivers.

[Spare Tire]

I got confirmation from eddie that the USB is 32-48kHz 16bit only:

USB output max at 32-48.1 khz only .
You might select 96khz at the soft wear ----It is just running 96khz at procceding and still output at 48.1 khz max.

That was my main gripe. Aside from that, it seems like the upsampling functions will only officially work with 44.1 to 88.2 and 48 to 96 respectively. Though i have found that 32kHz works with upsampling to 96 without noise, i think maybe 96kHz will work with any multiples of 8kHz, and 88.2kHz will work with any multiples of 11025Hz. I can't confirm for shure, but the math would work elegantly that way

Asked eddie if there was any way to mod the usb-to-spdif circuit to accept something better like up to 96/24. We'll see if there's any salvation for this unit. As it sounds quite good even as it's burning in right now, and it sure looks at lot better than the ugly DA7.2x

[he46570]

Thanks so much for your posts.

My E2.03e is arriving this week, and I should be able to give a different perspective to this device. I used to manufacture high-end tube equipment, so I should be able to look at the circuit and review its modding potential. I'll be in touch!

Cheers.

[Spare Tire]

It's been a couple of days and i haven't had an answer from eddie about the USB. I guess that's the taboo about this DAC. It's rated 192/24 but it never mentionned that the usb was only 32-48/16.

[he46570]

OK, folk. I received the DAC, and I'm thoroughly enjoying it. Unfortunately I don't have a 7.2X to compare it to, but it sounds great through my Etymotic ER4S phones. It does sound quite a bit better than my HPDAC, though - more detail, bass, far better highs, smoother and great dynamics.

I'm currently using the USB inputs with no issue. My laptop recognized the USB device immediately and self-installed the drivers (Windows XP). Interestingly this device is called "USB Audio CODEC" compared to the HPDAC which is called "USB Sound Card".

So rather than sit there and enjoy it - of course I decided to take it apart. The DAC is pretty easy to disassemble. The back plate comes off by removing all the rear screws (all hex keys, and the two sizes already come in the box). The body of the chassis is removed by unscrewing two screws on the top and two on below affixed to the front plate, and a few screws underneath attached to posts then attached to the PCBs and power transformer. The two pieces of the body of the chassis then slide rearward and come off.

The result is as pictured.



All in all I must say I'm very impressed with the construction quality and parts quality. ALl the major electrolytics are high-grade Nichicons. There are some smaller ones which may be tantalums, but they are unmarked. The power trannie is impressively large, and has a central epoxy core to reduce vibration. The power supply has a common mode input choke - a great idea - and all the rectifiers have suppressor caps on them.

A few things to note. One crystal was replaced with Eddie's TXCO. I haven't bothered to do an A/B comparison with what was in there before. The red modules appear to the the same ones in the 7.2X. The encased box with "Hi-Fi Audio System" printed on it appears to be the same one used in the EF3.01, though the EF3.01 has two of them. I tried to pry off the top, but couldn't... the screws were easy enough to remove, but it appears the case has been glued shut, or perhaps everything has been encased in epoxy. I'm assuming the "Hi-Fi Audio System" module is a discrete circuit for the headphone amp.

So what is there to improve in the DAC? I dunno, honestly. It looks so well put together, and parts quality is very high. There is an op-amp (OPA2604) just outside of the "Hi-Fi Audio System" module - I don't know what it's for, but presumably that can be swapped out for something else. But frankly I'm gonna leave this alone for now, and enjoy the music.

Let me know if you have any questions.

Robert

[Dave_M]

Hi, is there any CE marking or similar on the back?

Otherwise that DAC looks dangerously unsafe. There should be a dedicated earth wire to the case. It looks like the power on off switch passes mains voltages. If that fails you could be electrocuted. Each board should also be earthed as well in case the transformer fails.