[seeteeyou]

We do get both coax and Toslink from Oppo anyways, though the question might be more about how good the quality of coax inputs of non-Chord DACs would be when compared to the Toslink inputs of (jitter agnostic) Chord DACs.

Some DAC chips happen to convert DSD to PCM internally, while earlier products from Chord would also decimate DSD to PCM for processing.

For editing purposes, it's also easier (maybe necessary?) to convert something into PCM since that's really what editing software could work with.

---

BTW, it's inevitable that we're giving up M-Scaler whenever we're going for the Toslink route. However, what if connect one of the coaxial outputs to something like RME ADI-2 DAC?

http://www.rme-audio.de/en/products/adi_2-dac.php

The SPDIF input signal can even be recorded via USB - as one would expect from RME the DAC is a true 2 channel Full Duplex audio interface.

Granted we're getting only 176.4kHz instead of 705.6kHz but theoretically that's still 250,000 outta 1,000,000 taps. Once we've recorded the output from Blu2, we don't have to deal with the noise from Blu2 anymore while we could even bring those upscaled files with us and play them while we're away from home.

How about 500,000 taps? There's a company in Taiwan selling this CT7601 Audio USB Bridge for 200 bucks and its coax input could support up to 705.6kHz / 768kHz

http://www.comtrue-inc.com/index.php/products
http://www.comtrue-inc.com/index.php/downloads2/category/2-audio-usb-bridge

However, we're getting dual coax outputs from Blu2 instead of a single connection. I got in touch with their engineers and they informed me that CT7601 could only support one coax instead of both of them. In other words, the best we could do is 352.8kHz / 384kHz outta 705.6kHz/768kHz. A half million taps still ain't too bad, though.

Once we've saved those 352.8kHz / 384kHz lossless files to micro SDXC cards, we could also connect our DAPs to Hugo 2 / Mojo via USB since the coax outputs from many DAPs should be only good for 192kHz.

For recording the full blown 1,000,000 taps, most likely we'll have to wait for a future standalone M-Scaler with both USB input and output. I already asked them about combining both outputs from two separate CT7601 boards but coax connections could be somewhat tricky. Technically speaking we could indeed sync both boards with the same clock but in the end we might still end up with two "supposedly" identical files but the checksum might not necessarily match.

 

[Articnoise]

Like you, I would go with what sounds better even if it turns out what I am preferring is distortion but somehow, I'm not convinced that what I'm hearing is distortion and so the jury is out but I'm at least more aware of this possible bias.

There is a way to level the playing field between local playback on your Zenith SE and streaming from a NAS or from Tidal and that is with a very good network switch. The audiophile landscape will soon be littered with various network switch options from the likes of SOtM. Uptone Audio, and Aqvox. Last week, a company called The Linear Solution (based in Washington state), sent me their customized OCXO switch for trial and feedback. Essentially, this is a commercially available 8-port TP-Link switch that they "gutted" and improved with better capacitors, regulators and an OCXO clock that they sourced from Japan. To their knowledge, this is the only network switch that incorporates within the chassis of the switch an OCXO (other companies who offer an OCXO option for a switch house the clock in a separate chassis). They also designed a special linear-regulated PSU just for this switch that meets the 2.2A demand of the switch + OCXO and I'm told the combo of modified switch with PSU is available presently for an introductory price of $659.

http://thelinearsolution.com/ocxo_switch.html

I have several switches on hand to compare against including a stock Netgear GS108 that incorporates John Swenson's SMPS tweak, a Paul Pang TCXO network switch, and a SOtM-modified DLink 5-port switch clocked by an sCLK-EX/REF10 and powered by a 5V rail from a Paul Hynes SR7. At this time, I am using the SOtM-modified DLink switch as my reference and when fronting the Zenith SE, the improvement in less harshness, detail clarity and immediacy is very evident. I also have a Netgear cable modem / router modified by SOtM and so this cable modem / router is also being clocked by an sCLK-EX / REF10 and while this unit yields an obvious improvement, the improvement I get with the DLink switch is considerably more noticeable. With this DLink switch in place and connected to my Zenith SE via SOtM's dCBL-CAT7, if there is an advantage with local playback vs NAS playback, it is almost too small to tell and not of any great significance. With Tidal streaming, depending on the quality of the master used by Tidal, sometimes Tidal streaming sounds better.

This OCXO switch from The Linear Solution is at least as good as my DLink switch and with further burn-in, it may turn out to be convincingly better. At the present time, my DLink switch has a touch more smoothness and a touch less glare but music streamed through the OCXO switch definitely has more of a "you are there" presence and this switch seems to still be improving. If I can get this slight harshness to go away, my preference is definitely for The Linear Solution OCXO switch. Compared to "no switch," the Netgear GS108, or the Paul Pang TCXO, I would definitely go with The Linear Solution switch. When you factor in the cost of a REF10 and sCLK-EX board, my SOtM-modified DLink switch can only be considered a good value when these items are also used for other components such as a tX-USBultra or sMS-200ultra. While the Aqvox SE switch appears to be a better standalone value at 798 Euros (or nearly $1k USD), I have not yet heard this switch and so I can't really render an opinion but at only $659 for The Linear Solution switch, this switch will be tough to beat.

Within a few month's time, hopefully, we'll have a new switch from Uptone Audio and SOtM (which will be unique as this switch will include both RJ-45 and SFP ports) to evaluate but I am convinced that streaming from a NAS or from Tidal need not be inferior to local playback.

I have read some nice impressions of this switch lately and was thinking about maybe try it out my self - your post here made it tough to resist, but then we have the new switch from Uptone Audio and SOtM which complicate my decisions

Thanks for sharing your impressions and discoveries Roy!

 

[seeteeyou]

https://www.computeraudiophile.com/...ns-thread/?page=4&tab=comments#comment-798083

You might want to sit tight on the "audiophile Ethernet switch" thing for a few months. Were are working on something (yes a switch, but from the ground up, not a modification of anything) but we are not ready to discuss the details. As usual for UpTone, it will be radically different, very effective, and a great value.

The second half of 2018 is looking better than better, though we might wanna find something that's even better than JCAT NET Card FEMTO.

 

[Whazzzup]

man dsd 128 on antipodes is real nice.

 

[romaz]

Please kindly correct me if I were mistaken, I don't see a significant difference between the digital section of UDP-203 versus that of UDP-205. If that were the case, those of us who are looking for an I2S output should have lots of fun with I2S mods available from either Dr. Lee or JVB Digital

http://pcaudio.tistory.com/546
http://www.jvbdigital.nl/jvb.asp?page=vanity203
http://www.audiopraise.com/forum/read.php?13,838
https://www.oppostore.nl/vanity-hd-for-oppo-udp-203.html
http://www.jvbdigital.nl/jvb.asp?cur=2&level=sdi&page=front
https://forum.psaudio.com/t/transforming-vanityhd-card-for-oppo-203-into-usable-i2s/4427
https://static.webshopapp.com/shops/170546/files/175162763/vanity203-user-manual-v2.pdf#page=9

It's also worth mentioning that we could even get UDP-203 with Vanity203HD, then connect its quad coax outputs to four units of Hugo 2 (or even DAVE) for 5.1 / 7.1-channel materials.

As if that Korean OCXO (27MHz SBtron SBOC25) were not good enough, we could also reclock the I2S signal with even better clocks

http://www.diyaudio.com/forums/group-buys/207438-ian-asynchronous-i2s-pdif-fifo-kit-buy.html
https://www.computeraudiophile.com/...chronous-i2s-fifo-projectquot-on-diyaudiocom/

Then convert I2S to either coax or Toslink afterwards

https://github.com/iancanada/DocumentDownload/tree/master/FIFO II series/SPDIFboard

BTW, there are even better 45.1584MHz and 49.152MHz clocks for Ian's FIFO reclocker such as Pulsar Clock

http://www.pulsarclock.com/Clock.html
https://www.amb.org/forum/pulsar-clock-t3477.html
http://www.diyaudio.com/forums/vendor-s-bazaar/258789-pulsar-clock-ultra-low-noise-ocxo.html
https://www.head-fi.org/threads/lks-audio-mh-da003.745032/page-20#post_12775721

Manuf	Name	Type	Fc(MHz)	BW(Hz)	RMS Jitter (fs)
Abracon	ABLNO	XO	100.00	10.00-10.00k	200.923
Crystek	CCHD-575	XO	100.00	10.00-10.00k	151.769
Crystek	CCHD-950	XO	100.00	10.00-10.00k	226.627
Accusilicon	AS318-B	XO	100.00	10.00-10.00k	200.494
Vectron	OX-405	IHR	100.00	10.00-10.00k	89.534
Vectron	EX-401	OCXO	100.00	10.00-10.00k	89.658
Micro Crystal	SCOCXOH	OCXO	100.00	10.00-10.00k	159.949
Dynamic Engineers Inc	OCXO3308C	IHR	100.00	10.00-10.00k	69.452
CTS	VFOV504	IHR	100.00	10.00-10.00k	89.620
MTI-Milliren	221	OCXO	100.00	10.00-10.00k	28.356
Pulsar	Pulsar Max	IHR	100.00	10.00-10.00k	35.115
Pulsar	Pulsar Typ	IHR	100.00	10.00-10.00k	20.174
Morion	MV269	OCXO	100.00	10.00-10.00k	450.651
Magic Xtal	MXO37H/14	IHR	100.00	10.00-10.00k	58.573
Magic Xtal	MXODR	2xIHR	100.00	10.00-10.00k	25.974

And then there's also NDK DuCULoN

http://www.ndk.com/en/news/2015/1190895_e.pdf
http://www.ndk.com/en/ad/2013/001/pdf/c_NH47M47LA_e.pdf
http://www.diyaudio.com/forums/digital-line-level/314998-ndk-duculon-diy.html
http://www.diyaudio.com/forums/digi...mate-weapon-fight-jitter-360.html#post4372663
http://www.diyaudio.com/forums/digi...mate-weapon-fight-jitter-402.html#post4600591

Manuf	Name	Type	Fc (MHz)	BW (Hz)	RMS Jitter (fs)
Abracon*	AOCJY1	OCXO	10.00	10.00-10.00k	1596.037
Symmetricom	SA45s CSAC	CSAC	10.00	10.00-10.00k	1428.198
NDK	NZ2520SD	XO	22.58	10.00-10.00k	62.696
NDK	Duculon	OCXO	49.15	10.00-10.00k	20.044
Crystek	CCHD-575	XO	22.58	10.00-10.00k	221.180
Crystek	CCHD-957	XO	22.58	10.00-10.00k	310.181
MSB	Galaxy	OCXO	24.58	10.00-10.00k	16.176

You're correct, Frankie. If all you want to use is the digital section, the less expensive BDP-203 should fit the bill and save you some money. Paired with the JVB SPDIF board, it may be even better, however, the JVB board uses its own clock and the quality of this clock is uncertain. Furthermore, by the time you purchase the JVB SPDIF board and pay the extra $200 for Oppomod's "Complete mod" for the BDP-203, you would have spent more money than upgrading a BDP-205.

As you've pointed out, the parts used for this upgrade could be better which is why I was a bit surprised by how much it improved. As always, it's how these parts are implemented that can make a bigger difference and combined with very short signal paths (i.e. no convoluted signal paths or long clock cables were necessary), the quality of the clock signal actually seen by the system board is possibly higher than the quality of the clock signal seen by components such as a tX-USBultra or sMS-200ultra even though they use superior clocks.

One of the things I learned about the REF10 is that Mutec's published phase noise measurements were taken from the BNC output and not from the clock itself and so comparing the phase noise plots of REF10 to another external master clock may not be a true apples to apples comparison and this highlights my point. I predict that by the time the clock signal goes through whatever length clock cable you use to connect the REF10 to something like the sCLK-EX board in my tX-USBultra and by the time that clock signal reaches the USB hub in my tX-USBultra, that signal may have become significantly degraded and so this is the potential benefit of physically placing a good clock in the same location as the old clock.

For those wanting to take advantage of the feature set of the Oppo but also wanting to take the Oppo much further than I have, there is certainly the option of approaching someone like Paul Hynes or Sean Jacobs for a custom external 2-rail LPSU although I would specify that the final regulator board be placed in the chassis to keep impedance to a minimum. Obviously sourcing a better clock should result in even better results taking into account the potential caveats that I mentioned above.

 

[romaz]

Way to go out on a high note, Roy! Yet another game-changer, and an amazing find.

I agree with you, that if one can live with the Oppo as a DAC, after the PSU and OCXO clock improvements, this is an incredible one-box solution. But if you want to use an external DAC, I do think the Oppo's limitation to the Toslink output (i.e. no USB output) makes this a rather narrow solution. Yes, for you Chord owners, because Toslink is that family of DACs' best input, this is a great solution, but most DACs have rather poor Toslink input performance, so for these DACs, the trifecta will have better performance.

One other caveat - the Toslink output also limits one to PCM-only, and 24/192 (or is it 24/96?) and under. Which is OK for a majority of people, but some of us have a lot of DSD and DXD content too, which we'd rather not transcode to lower-resolutions.

First-world problems!

Thanks, Rajiv. Yes, for someone like you who owns hundreds of DSD tracks, a Toslink connection presents its challenges but Toslink completely fixes a problem that can otherwise cost thousands of dollars to address and that is RF noise. You look at the lengths that server companies go to in order to keep RF noise that enters the ground plane to a minimum and these lengths often result in servers that costs thousands of dollars and yet in one fell swoop, with an inexpensive Toslink connection, you completely eliminate it. I believe this is one big reason why this Oppo unit performs so well in my system.

Toslink certainly has enough bandwidth to pass DSD64 without truncating data and so this 24/192 bandwidth limitation impacts only DSD128 and higher and of course, it impacts DXD. If you have a large library of natively recorded DXD or multi-rate DSD, then you're pretty much stuck with USB since AES/EBU and digital coax are hampered by the same limitations. The problem with the majority of DSD files, which we have discussed privately, is that natively recorded multi-rate DSD (DSD 128 and higher) are rare meaning that the large majority of DSD files you can buy today were edited, mixed, and mastered in PCM and so I'm not sure what you lose by transcoding back to the original roots of your DSD files unless your DAC simply doesn't sound that good playing back PCM files. In this situation, you will need to weigh the positives of completely eliminating RF noise in the ground plane vs the negatives of playing a format that your DAC doesn't excel in.

Aside from bandwidth, the problem with Toslink is that it is a high jitter connection and is generally avoided by most DACs and this avoidance of Toslink highlights something interesting. Most DAC companies these days claim that their DACs are immune to jitter when in fact, they are probably not and so it would be necessary to test and compare your DAC's inputs to see how Toslink fares. It's possible this Oppo solution is of greatest benefit primarily to owners of Chord DACs.

 

[romaz]

My 2 questions:

1. Would you expect that replacing both mR and IR with the Oppo would be a significant upgrade?
2. Would adding a Linear Solution switch into this direct chain further improve the sound?

The Oppo is physically too big to fit with my main rig downsizing aspirations, but I see an opportunity to kill 2 birds with 1 stone by using the Oppo as both A/V hub (Tier 2 sound qualiy) and my Tier 1 DAVE-based main rig. I was already considering the Oppo to replace my ancient DVD player in my A/V system. I'm still not going to spend any significant money until my SR7 arrives to give me a new reference point. Sadly, Paul has already missed his new delivery schedule - As 7th in queue in January, I should have had notice of shipment by now. Sigh....

BTW, Linear Solutions seem to have lots of interesting OXCO-based products, including LAN and USB cards and a wireless router.

BTW 2, whiist not spending much money on my main rig, I've had a very worthwhile incremental SQ boost by upgrading my Fidelizer Pro from 8.0 to 8.1 (still alongside Process Lasso). I've had mixed feelings over FP over the years, sometimes wondering just how much benefit it was providing. But with 8.0 the designer made a concerted effort to keep audio related processes within specific cores etc. Also, on his recommendation, I switched off hyperthreading. That seemed a nice incremental improvement, but now that he's just cleared that big release's snagging list in 8.1, well, that has been the most obvious improvement I've yet heard from any FP release. Highly recommended (alongside PL) for those who have basic Windows Servers and don't want to go the more extreme Audiophile Optimizer route. This further reinforces my view that improving the server still matters despite all the little black boxes downstream.

Regarding your 1st question, as you use a Chord DAVE, this modified Oppo connected to your DAVE via your Mapleshades optical cable (which is what I use) will probably represent a significant upgrade over your mR and IR. I say "probably" because I can't account for personal preferences but having owned both an mR and IR and even with both of these items powered by an SR7, they did not come close to performing anywhere as well as my combination of Zenith SE followed by a tX-USBultra whereas this Oppo does.

Regarding your 2nd question, the answer is emphatically yes. It was with connecting the mR to my server via a bridged LAN configuration where I found a good switch to be especially impactful. When placed before the Zenith SE and before this modified Oppo, the impact remains very significant.

Yes, Linear Solutions will be coming out with some cutting edge products later this year including a very unique server that incorporates a sophisticated 9-rail LPSU, 3 OCXO clocks (system, Ethernet, and USB) without resorting to PCIe cards and a very lightweight OS that runs completely from RAM with near zero latency and all of this will fit in a single chassis.

Regarding the SQ boost you are getting with things like AO, Process Lasso, and FP, one of the things I continue to prefer with my custom server over my Zenith SE is the ability to tune or tweak my sound signature to my preferences via software without having to resort to DSP or other processes that require considerable CPU horsepower.

 

[seeteeyou]

Lo and behold, here's the latest from SOtM

https://www.computeraudiophile.com/...reaming/?page=295&tab=comments#comment-798252

Hi, I would like to share the pictures how I've setup the sCLK-EX for UDP-203 and UDP-205, also I've built up the new power board for the digital input for UDP-205. The modification have been done impressively and if anyone would like to try out this modification, please email for more details and final quote.

That really didn't take long, though we might also think about modifying another Roon Endpoint with even better implementation of clocks for the Toslink output.

For instance, it's a good idea to dedicate one clock for multiples of 44.1kHz and then another one for multiples of 48kHz like this particular example below

https://community.roonlabs.com/t/pi-2-design-502dac-pro-audio-shield/27954
http://www.pi2design.com/502dac.html

Toslink and other outputs are also getting clean power separately from an ultra low noise high PSSR LDO linear regulator LT3042, then it also doesn't hurt to solder the (OCC silver with super large gauge) DC power cable directly to the PCB since we'd like to bypass that DC barrel with high impedance. Of course it's also necessary to replace both NDK NZ2520SD clocks with much better ones, though the question is where do we find the right cables for connecting the clock output of an OCXO to the clock input of the Roon Endpoint.

How about placing an OCXO right on top of the clock input like this? Though it's only good for DIL14 sockets

https://www.head-fi.org/threads/lks-audio-mh-da003.745032/page-20#post-12769955
https://www.head-fi.org/threads/lks-audio-mh-da003.745032/page-22#post-12827379

 

[Whazzzup]

Crap are you guys building servers? Is this what this thread is? Modding servers or building. Is there a thread for stand alone music servers?

 

[bmichels]

Sorry if this has been discussed before here, but I jump back into this thread without the courage to read the past 80 pages:

I recently realized that the MELCO has 2 ethernet ports: one to connect to the Internet and one to connect to the streaming DAC. I find this brillant since the Streaming DAC do not directly connect to the router and therefore the data from the Melco HD directly go to the Streaming DAC without going through noisy LAN/Switches. Plus, Melco pretend that the MELCO server also act as an Ethernet filter for data coming from internet or the LAN (while streaming from internet or a NAS).

- Is this what people here call Ethernet Bridging ?

- Does this theorical advantage really translate in better (cleaner) sound ? Does using a MELCO sound better than a Sylogic NAS ?

- my problem is that Melco cannot act as a ROON server, so if we want to use Roon, then we partially loose the beauty of this direct connection and we need another server or PC .
--> So... are there other dual port server (with the same direct connection to the DAC) than can run ROON Core ?

thanks for the answers, and again sorry to ask something that may be has already be answered here before.

 

[Fredoo]

--> So... are there other dual port server (with the same direct connection to the DAC) than can run ROON Core ?

Hi, you can contact The Linear Solution (please see post 1177 just above).
Cheers, Fred

 

[bmichels]

Hi, you can contact The Linear Solution (please see post 1177 just above).
Cheers, Fred

Thanks. Do you have a link to their webSite ? all I can find is " linear-solutions.com " . which seems to be another company...

Isn't Innuos Zenith server also providing this dual ethernet capability ? it is not 100% clear what those 2 ethernet ports are for ? and does the Zenith act as an Ethernet filter like the MELCO does ?

 

[Fredoo]

Thanks. Do you have a link to their webSite ? all I can find is " linear-solutions.com " . which seems to be another company...

Isn't Innuos Zenith server also providing this dual ethernet capability ? it is not 100% clear what those 2 ethernet ports are for ? and does the Zenith act as an Ethernet filter like the MELCO does ?

Hi, on thelinearsolution.com you’ll find contact information.

 

[bmichels]

thanks a lot. I have sent them a mail

 

[shyamwanne]

I have been in communication with Adrian at thelinearsolution.com about their new server coming out.

Their music server and endpoint (renderer/player) will be a very limited run released in august or September this year. An info page should be up this weekend on their website for pre-sale.

The operating system is their own proprietary OS “DreamOS” base on Linux kernel, customized to audio playback only. The Size of DreamOS is below 5GB And the Latency is AVG at +-50us. It can use Roon, HQPlayer, MPD or UPNP Mini-Server. Adrian said it is the most advanced OS and Sound out there, and should be better than anything out there at this time-line.

Pretty exciting