Cheapest way to enhance SQ with digital streaming


I could not stop but post my observations on remarkable SQ improvement with just a small change. I have been using digital streaming predominantly and have tried many things (cables, conditioner, room treatment). Everything does matter and they all did improve the quality without doubt and are definite investment. But after going through some reviews of fiber optic for Ethernet and their noise reducing properties compared to copper cables, I thought of giving a try. I connected Ethernet cable from router/Ethernet switch to media converter and fiber optic from this media converter (need SFP module) to second media converter (with another SFP module). Then regular Ethernet cable went from this second media converter to server/streamer. Basically the idea is to add fiber optic cable between the router and streamer, which will reduce the noise. But, wow, the difference was huge and amazing. Not sure how to explain the improvement, but the SQ was more fluid, melodious, fast, clean and separation of instruments were clear. The total cost for this was less than 150 (on Amazon). Apparently, as the fiber optic cable conductors are made of glass, they carry less noise compared to copper and also cost very cheap. Compared to money spent on other component in my system (have spent quite a lot), this is the most cost efficient and gave more improvement in the SQ. I am still amazed by this and very happy with what I heard. I feel negligent not to share with my fellow audiophiles, who are trying to get the best out and improve their system, sometime spending quite a bit for small difference. I guess, since the cost of trying this is so cheap (please don’t make me feel guilty for telling 150$ is cheap), try yourself and see the difference. On a different note, have been reading that use of Ethernet switch also does some “saucery” (I am no technical guy, but do read quite a bit of articles on audio) to reduce this noise to improve SQ and combination of fiber optic and the Ethernet switch takes SQ to different level. Will try some audiophile Ethernet switch sometime (may be uptone ethergen or Sotm snh-10g) to try out as I am very intrigued. 
romney80

Showing 28 responses by sns

I'm moving on from generic fmc and Sotm sms200 neo to the Sonore Opticalrendu, replaces one FMC and it's lps and sotm. Next will be optical out server such as Small Green Computer Sonictransporter I9 or diy solution with optical out, this gets rid of second fmc and it's lps, and upgrades present server. At this point all optical post server. At some point I'll get router with optical capability, lps on this, no switch. My goal is all optical post modem.

 

Full optical is way to go!

1. Yes, requires two fiber media converters, two ethernet cables,one into first fmc, one out into bluesound.

2. Shorter is better.

3. Lengths your talking about, no.

Finally had chance for first listen with opticalrendu running off Uptone JS2 LPS. Don't like hyperbole, but this is major upgrade over two TPLInk MC220 with Sotm SMS200 neo (all with LPS). Certainly one of best digital upgrades I've experienced, another level of resolution together with more relaxed, analog like sound. Easily worth the extra cost over the generic FMC.

Yes, ethernet vs optical cable cost is important variable to consider. Prior to optical I tried various classes of generic ethernet, worked up through Supra, Wireworld, and various AQ from Cinnamon on up to Vodka. Vodka substantially better from generic, cost becomes consequential when you add up all the individual runs. Optical cable cheap, and more importantly, higher resolving than any ethernet cable due to complete isolation. Per Romney80, noise free cable has a different sound, it is the absent of sound, you're only hearing pure signal.

 

The only thing that can alter the sound of cable is the converters and power supplies used with them, whereas ethernet cables have sound of their own. Ethernet cables have various shielding schemes and metallurgy, all effect the sound. I'd rather have the zero variable of optical cable.

@car123 Try lps on second fmc, this should be worthwhile improvement. Don't need to spend a lot of money on this, Teradak or something similar fine.Not surprised didn't hear much difference with fmc in front of server. Adding switch and fmc after Aurrender would make more difference. But then this requires adding usb renderer.

If one wants to both reduce clutter and optimize sound quality, two ways to go. FMC after server, this requires two computer setup, aka separate server and streamer, both optical enabled. The second is one computer setup, optimized usb out of server. With optimized usb out of server, add on filters shouldn't be needed.  This assuming one using usb dac.

I'll add if your intention is best sound quality, perhaps more clutter provides better sq. Simpler may be betterr, or it may not. I see so many positive reviews on so many setups, if someone has proof of the definitive best setup I'd like to hear it. I'm all for experimentation and reporting results, adds to knowledge base.

For those using usb out of servers direct to dac, many have found various usb filters, reclockers and better usb cables are nice upgrades. Usb ports direct off motherboards are perhaps weakest link in most music servers.

 

I'd go this route vs. optical network conversion in front of server. Optical conversion is most likely to work in network AFTER server. Servers are noise makers, the reason why usb filters and optical conversions after server are so effective and popular. Cleaning network prior to server only to reinstate noise produced by server is somewhat ineffective and illogical. I've even found it somewhat ineffective after treatment post server, one audiophile switch I tried resulted in diminished sound quality. On the other hand, adding quality lps to my router improved sound quality, so mixed results here.

You'll find diverse opinions on most streaming products, in the end one has to determine for self. The good thing about the FMC experiment is it can be done for less than $100, where can you find such a deal in audio! If experience promising you can then optimize with linear power supplies, from there you can go to even further optimization with product like Sonore Opticalrendu.

The ethernet route has its own set of basic equipment and further optimization.

@tksteingraber Not sure what you mean by after streamer. Streamers generally do usb straight into dac, if you converted usb to fiber you'd have to have fiber input on dac. Generally, fiber is used to convert ethernet to fiber, this goes into server/streamer or separate server or separate streamer.

 

Now if you're talking server, you can do FMC AFTER server ONLY by using separate streamer (examples my Opticalrendu) which has fiber input, usb output to dac. Without separate streamer FMC can only be done UPSTREAM of server/streamer.

 

I think people get mixed up by all the different words applied to streaming equipment. Servers are most common of streaming equipment, vast majority have streamer built in, if one is using usb or another port direct into dac, this has built in streamer. Streamers are a separate computer that is placed downstream of server and just prior to dac, steamers also called Roon endpoints, usb renderers. Streamers have their own  operating system since they're computer, this allows them to communicate with the Roon Core installed in server, why they're called Roon Endpoint, they also convert ethernet or optical to usb, reason they're called usb renderer.

 

Three theoretical advantages to separate streamers, one is removing Roon endpoint duties from server, means less processing by computer within server which means less noise produced.  Two is ability to do FMC after server, noise produced by server can be completely eliminated. Three, the usb rendering process is generally filtering and/or reclocking of usb creating improved usb rendering vs. what may be untreated usb coming directly out of motherboard on server, result is better feed to dac, cleaner, less jitter.

 So, one has option of keeping it simple, using general service computer or computer optimized for music, this is called the server. take usb straight to dac. The other way is more complex as I've tried to explain above.

@dbphd  You got it. Streamers take ethernet or optical input (in rarer cases), use various outputs to dacs. Streamers can also be built into dacs, like your Ayre in which case they are streamer dac. Your ultrarendu is also streamer. If using usb output on any streamer its also called usb renderer, if using Roon also called Roon endpoint.

As mentioned previously, a server can also be streamer, although simply called server, all servers connecting to dacs are inherently servers with streamer capability.

Generic FMC without lps good start, get lps on second fmc next move, lps on first last move. Opticalrendu would be most effective if rest of streaming setup top notch, if dac is weak link replace that prior. Once all this in order Opticalrendu can bring substantial improvements to streaming sound quality, remember to calculate for quality lps as well, there is good reason why Signature rendu SE commands premium price!

@tksteingraber Did you try FMC set between macbook and dac with FMC set between router and macbook removed?

 

I'd look to dedicated server vs macbook for next upgrade.

 

Good to see more of these comparisons. In future we need to see comparisons of optimized ethernet vs optimized optical setups in same audio system, not sure I've seen that.

 

At this point, many iterations of ethernet devices, with optical Sonore only device I'm aware of.

When doing the cheap fiber mod, using  stock wall warts isn't extracting full value of mod. LPS on second fmc is next step, I found lps on first fmc also helps just a bit.

With ethernet solutions, cable quality matters.

@goofyfoot I'd rather run longer optical than ethernet cable, less signal loss, optical also immune to emi/rfi, so first FMC close to modem, second close to Pi. Power at least the second with lps.

@audphile1 Even with FMC, ethernet cable quality still matters, your experience illustrates that. Using your best ethernet cable closest to dac makes sense. Upgrade all your ethernet and be prepared for further gains. Upgrading from Supra and various other ethernet to AQ Vodka was well worth it for me.

 

Further gains with generic FMC can be had with even better lps, power cords, and connection to power conditioners.

While I hate to say this, as generic FMC can be very nice upgrade, one can still further optimize optical. Generic FMC use pretty lousy clocks and noisy regulators, last FMC in chain injects it's own noise into network, and this, most likely close to dac which is worst place to inject noise. So, we replace last FMC with Sonore Optical Module, much better Femto clock and four low noise linear regulators, less noise, better timing for less jitter. I have one coming supposedly in July (chip issues), I expect good things.

 

Just goes to show for every upgrade in streaming, there is yet another upgrade. Streaming  and network optimization has yet to reach a static optimal solution.

Continuing experimentation with optical continues causing me to reassess its value in all instances. I've been using generic FMC, with lps and quality power cables prior to server for past month or so, thought I had pretty good handle on sound quality. So went back to router with lps, nice power cord straight into server (optical remains post server), much better sound quality sans FMC. This back to original configuration, always like to do A>B>A comparisons to check for confirmation bias.

 

IME, the two FMC, while having slightly lower noise floor, in the sense I could hear lowest level info slightly easier. it was at expense of more analytical presentation. More hifi than sense of realism I get without FMC. Transparency really took a hit with the FMC, more closed in, veiled sound quality, Decreased sound stage size, less image dimensionality, less air around performers, all contributed to less sense of performers in room.

 

My take is generic FMC may take more than give. Less than optimal implementation of optical for reasons given in my last post, added complexity with streaming chain may also be liability.

@esarhaddon Good catch, completely forgot about that. Who makes the good polished plastic and/or glass optical cables? I'd like to try this.

@esarhaddon I don't run Toslink, only network optical cables. I'm using optical cables  Small Green Computer sells with Sonore equipment, assume they are polished.

 

I do recall reading something about differences in optical cables some years ago on Audiophilestyle forum, forgot about this until your post.

@tksteingraber The only reason lps on at least second FMC would be lack of resolving ability of entire audio system, or lps not of sufficient quality. I can hear differences with lps on only one or two FMC, with different lps,  with different power cords used with the lps, and with lps connected to battery storage device or ac power conditioner. Try a Teradak lps with decent power cord and report back.

@tksteingraber Well that's good, proves your system resolving enough. Don't know if I mentioned in this thread, but I have tried battery power with dacs and lps for various streaming devices, always went back to ac with my BPT 3.5 Sig. power conditioner as superior. I too hear warmer/softer with battery, but this less satisfactory with time, noticeable decrease in transients, micro dynamics, music lacked excitement, involvement with battery power.

 

My take is battery power could likely be preferable to non optimized ac, but with dedicated lines, top notch grounding and power conditioning, ac better.

@tksteingraber Sounds good, your results what I'd expect. Just a reminder, power cables on lps and ac feed, power conditioner will also help. Lps and same upgrades on other FMC will also help, although to lesser degree.

 

Even the dc cables (from lps to FMC) can have an effect, at this very moment another thread here in regard to this upgrade.

 

@mitch2 Perhaps huge and amazing may be over exuberance over an initial, formerly unheard difference, sure many of us have been here.

 

In going back in my mind over all streaming changes since my initial foray some seven, eight years back, there were times when a change brought about this same over exuberance. Over time I can now see changes have all been incremental, however, if I were to compare my original setup to present it would be huge and amazing difference. Perhaps you need to do this comparison in order to reaquaint yourself with the efficacy of your upgrades.

 

For me, the higher resolving my system, the easier it is to hear the smallest changes.

 

Based on your description of setup, weakest link I see is the 45' ethernet from switch to gigafoil. Moving Silent Bonn close to router, replacing standard switch would seem to be good move. I'm lucky in that my internet service comes into my dedicated listening room, but at one point had this setup at far end of room, running 25' ethernet from modem to router (which is close to audio system). I then decided to setup modem close to router via 25' longer coax, so now I could use 1m ethernet from modem to router, no ethernet longer than 1.5M in entire streaming setup. So between shorter ethernet which also meant I could use better ethernet (
various generic to Supra to AQ Cinnamon, long cables,  to AQ Vodka, short cables). Both shorter and better quality ethernet cables were meaningful upgrade for me.

Yep, its usually the case several incremental upgrades result in an overall large upgrade. I nearly always do A>B>A comparisons to confirm I'm hearing what I think I'm hearing, especially with streaming. I've done a fair amount of experimentation with streaming equipment, expectations have never been so tested, A>B>A results exposed some 'upgrades' as not being real upgrades, lateral or backwards move on my part.

This is why I've kept audio journal for twenty plus years, document every change. This has been extremely beneficial in moving forward rather than constant lateral or circular movement. Allows one to systematically build system with strong foundation, end of equipment churn.

I agree on switch front, but can we say this about all switches? Some report switches in $1k range and below are of little or no benefit, those priced much higher of great benefit.