Toslink switch and TosLink/Coaxial SPDIF converter - Review and measurements

[DrCWO]

I like to share some additional thoughts regarding the measurements with the ADI-2/4.

To measure Jitter you have to have an ADC which has a super stable reference clock. If you connect a DAC to it you can measure the jitter performance of the DAC.
In the first setup of @Rja4000 the situation is different as the ADC and the DAC of the ADI-2/4 share the same clock! This means: If there is a Jitter introduced by the DAC you can‘t measure it as the ADC jitters with the clock as the DAC. This is exactly the reason why you see absolutely no Jitter in your first measurement

Conclusion:
The Jitter performance of ADI-2/4 is great but not as great as shown here. To measure the introduced Jitter by the DAC you have to use two ADI devices. One for the ADC as reference and one for the DAC under test.

 

[Rja4000]

The Jitter performance of ADI-2/4 is great but not as great as shown here. To measure the introduced Jitter by the DAC you have to use two ADI devices. One for the ADC as reference and one for the DAC under test.

Well,
it makes no difference

Here is another set of measurements, done quickly.
DAC is still the RME ADI-2/4 Pro SE but the ADC is now an RME ADI-2 Pro fs (AK4490 edition).

Setup is
PC > USB > RME ADI-2 Pro fs (AK4490 generation) > Toslink [ > TosLink Switch > TosLink] > RME ADI-2/4 Pro SE (as a DAC) > XLR > RME ADI-2 pro fs (as an ADC) > USB > PC

SINAD 192kHz

Without TOS Link Switch

Figure is lower than above because ADC is less performant

with TosLink switch

Jitter measurement 48kHz

Without TosLink switch

With TosLink switch

Still very boring, IMO.

 

[DrCWO]

No, you should try this:
usb -> ADI-2/4 -> Toslink -> ADI-2 DAC analog out -> ADI-2/4 analog in -> usb

So ADI-2/4 is reference and ADI-2 the device under test

Edit:
Hi @Rja4000, did not read you post exactly, sorry. My proposal is exactly what you did just with exchanged ADI-2 and ADI-2/4.
Result is stunning. Well done @MC_RME

 

[Rja4000]

Ok, now that you recognize it's not needed to just swap the 2 devices, I'll admit it:
I checked it the other way around too


DAC is now the RME ADI-2 Pro fs (AK4490 edition) and ADC is the RME ADI-2/4 Pro SE .

SINAD 192kHz

Without TosLink switch
The SINAD of this old generation AK4490 ADI-2 Pro fs is pretty respectable, still.
(Remember, it's a DAC + ADC measurement - Mono mode - volume -2.5dB on the DAC)

With TOSLink Switch

So...
Even the old gen RME ADI-2 Pro fs doesn't give a sweat at this.

I think the demonstration is (over) done now.

 

[Rja4000]

---

Would that be more clear if I add those drawings to the reviews ?

 

[MC_RME]

To make it fully clear to everyone you might have to add 'Clock Master' into the blue field and 'Clock Slave' into the yellow one.

 

[OldManMatt]

I bought this exact toslink/coaxial switch just a few days ago..and I'm actually using a RME ADI-2 FS with it. I'm using the Switch for my TV optical to coaxial so I could use the optical out of a WiiM mini to the RME.

I definitely had concerns about added noise from the Switch. I listen to a lot of music and music videos from the TV and I find the quality of sound to be quite good. For whatever reason, at times, I kinda prefer the sound from my TV to RME over "Hi-Res" streamers to RME.

Subjectively speaking I felt the quality of sound through the Switch was good. I didn't feel like I was losing or adding anything extra to the signal. Obviously this can be hard to determine without any data or proper comparison test..so this review helps me feel a little better about adding this unit into the mix. If it is degrading the signal in any way I didn't notice. To be honest I wasn't sure if it would be a keeper, but it was cheap enough and Amazon is great about returns.

 

[Rja4000]

Recommended Optical Switch to use with ADI-2 DAC (Page 1) — ADI-2 & 2/4 Pro series, ADI-2 DAC series — RME User Forum

I purchased one on september 11th.
No news since then...

EDIT: Got news today.
It's on its way.
Too many orders since 2 weeks it seems ;-)

EDIT2: Busy time for me at work (Q4 rush, as every year) and at home (Preparing next photo exhibition).
But I received the device and longer FOs.
So when time will allow, I'll test both.

 

[JamesYeomans]

I purchased one on september 11th.
Now news since then...

They can take a while as he makes them to order. Took a couple of weeks for the two RCA ones to arrive in the UK. I'd give him a nudge if you've heard nothing
The BNC variants take much longer as he batches them up. Took about 6 months for my BNC one to be made & shipped to me.

 

[MCH]

I purchased one on september 11th.
Now news since then...

if in the meantime you fancy diying one yourself, this guy at diyaudio made one that i find better, with display. It is based on the same wm8805 so i bet the performance will be the same:

Whazon, a S/PDIF switch built around WM8805

A S/PDIF switch can be useful when we have more digital audio sources than the number of inputs available on a DAC. I wanted to do this a long while ago and finally managed to get at it recently. This little project makes use of the internal 8-channel selector offered by WM8805 S/PDIF...

www.diyaudio.com

 

[gumblag]

I measure-checked 2 digital Toslink devices :
Toslink switch + Bi-directional TosLink/Coaxial SPDIF converter

View attachment 310509

I needed to extend my digital inputs capacity for home living room and another location.
So I purchased
- a 3 inputs - 1 output small TosLink switch - with remote. (30€ - Audiophonics)
- a bi-directional Coaxial/TosLink SPDIF converter (26€ - Amazon.fr)

Both are pure digital SPDIF devices.
No DAC or ADC here.

As you may see on the picture, those are VERY small devices.
Even the remote is tiny, compared to the XLR connector that holds it straight.


View attachment 310511

The TosLink switch has 3 inputs and 1 output.
Input may be selected through the remote or the front switch.

The Coaxial/TosLink SPDIF converter has a coaxial SPDIF (RCA connector) and a Toslink on both sides: Inputs on one side, outputs on the other side.
A switch allows to select which of the 2 inputs you play from.
Both outputs are then receiving the signal (More on that later)

Both devices are powered with 5V and come with corrsponding USB connector.
The convertor came with a 5V adaptor, but with a US plug, so not really usable for me.



I wanted to validate those devices for transparency.

So I've set-up some measurements using a RME UCX II (digital source) and the RME ADI-2/4 Pro SE.
As Amir is doing, I measured performance on the analog side, after the DAC receives the digital signal through the device I wanted to test.
If they are transparent, we should see no difference in measurements.



1. Testing with the RME ADI-2/4 Pro SE as a DAC

First, I've performed measurement with my ussual tool, the RME ADI-2/4 Pro SE as a DAC
I measured the analog loopback (DAC + ADC) of the RME ADI-2/4 Pro SE when fed with a digital signal coming through those devices.

Warning: This will be boring !
So you may want to jump to the next paragraph directly


SINAD

Because I wanted to check the functionality, all SINAD tests here were perfomed at 192kHz.

First, through the TOSLink switch
NB: I tested the 3 inputs, and there is no difference

View attachment 310514


Then through the Coaxial>TosLink SPDIF converter

View attachment 310515

Then the same converter but in the other direction: TosLink > Coaxial SPDIF Converter

View attachment 310517


And, in case you were wondering, the RME ADI-2/4 with SPDIF input (analog loopback)

View attachment 310518



Jitter
To save some time, I cascaded the 2 devices:

> SPDIF Coax > Coaxial/TosLink SPDIF converter > TosLink > TosLink Switch > TosLink > RME ADI-2/4

View attachment 310521

then
> TosLink > TosLink Switch > TosLink > Coaxial/TosLink SPDIF converter > Coaxial SPDIF > RME ADI-2/4

View attachment 310522


Multitone

Multitone tests were performed at 192kHz.

Same here
> Coaxial SPDIF > Coaxial/TosLink SPDIF converter > TosLink > TosLink Switch > TosLink > RME ADI-2/4

View attachment 310585

Then
> TosLink > TosLink Switch > TosLink > Caoxial/TosLink SPDIF converter > Coaxial SPDIF > RME ADI-2/4

View attachment 310586

Just perfect.

I told you this would be boring





2. Testing with the Topping E30 II Lite as a DAC

I've been challenged on the fact that the RME is almost a "perfect" tool.
So there is a risk the test is not relevant, since the RME would fix any issue with the incoming digital signal.

That looked like a valid concern.

So I took a DAC I had at hand that I knew had some issue with Jitter on SPDIF input:
The Topping E30 II Lite I reviewed here.

So... here we go again.



This time I started with
Jitter measurement


Direct input in the DAC

RME ADI-2/4 Pro SE optical output > TosLink > Topping E30 II Lite > Analog output (L) > RME ADI-2/4 Pro SE

As you may see in the Topping E30 II Lite review, Jitter performance on SPDIF has some issue.
Not catastrophic;, but not stellar either.


First, for reference, with none of the devices we want to test in the loop.

Here, I kept the measurement of the highest secundary peak, to allow some comparison.
NB: This figure doesn't mean much by itself - don't compare it to a SINAD value !

View attachment 310591


Through the Coaxial/TosLink SPDIF converter

> Coaxial SPDIF > Coaxial/TosLink SPDIF converter > TosLink > Topping E30 II Lite > Analog output (L) > RME ADI-2/4


View attachment 310592

Here, we see a difference ! And this is not good.
The second peak (and all other peaks) raised by almost 12dB.

Is that a concern ? I don't know.


Let's try the other direction
(More important, since that's the one I, personnaly, plan to use )


> TosLink > Coaxial/TosLink SPDIF converter > Coaxial SPDIF > Topping E30 II Lite > Analog output (L) > RME ADI-2/4

View attachment 310593

In that direction, we have less degradation of the result.
At least that's good.
It's still an average result, but it didn't become much worse than what the DAC can do.


Through the TosLink switch

> TosLink > TosLink Switch > TosLink > Topping E30 II Lite > Analog output (L) > RME ADI-2/4

View attachment 310594

Same here.
Not worse than what the DAC can do.


SINAD measurement

Let's check the impact of this on SINAD, if any.
(As above, SINAD measurements here are at 192kHz)

Direct input in the DAC

RME ADI-2/4 Pro SE optical output > TosLink > Topping E30 II Lite > Analog output (L) > RME ADI-2/4 Pro SE

View attachment 310573
Our reference here.


Through the Coaxial/TosLink SPDIF converter

> Coaxial SPDIF > Coaxial/TosLink SPDIF converter > TosLink > Topping E30 II Lite > Analog output (L) > RME ADI-2/4

View attachment 310574

Not much impact.

Let's try the other direction

> TosLink > Coaxial/TosLink SPDIF converter > Coaxial SPDIF > Topping E30 II Lite > Analog output (L) > RME ADI-2/4

View attachment 310575

A bit more distortion, maybe. Still excellent.

Through the TosLink switch

View attachment 310576

We see again a slight decrease in performance.
That remains very good, though.

Interestingly, the performance here degrades in the opposite direction than the Jitter above.
Why ? No clue.


Multitone measurement

Direct input in the DAC

RME ADI-2/4 Pro SE optical output > TosLink > Topping E30 II Lite > Analog output (L) > RME ADI-2/4 Pro SE

View attachment 310577

That's our reference.


Through the Coaxial/TosLink SPDIF converter

> Coaxial SPDIF > Coaxial/TosLink SPDIF converter > TosLink > Topping E30 II Lite > Analog output (L) > RME ADI-2/4

View attachment 310578

Same thing: a bit more noise, but result remains excellent


The other direction

> TosLink > Coaxial/TosLink SPDIF converter > Coaxial SPDIF > Topping E30 II Lite > Analog output (L) > RME ADI-2/4

View attachment 310579

Nothing to complain about


Through the TosLink switch

> TosLink > TosLink Switch > TosLink > Topping E30 II Lite > Analog output (L) > RME ADI-2/4

View attachment 310580

A bit degraded still, but anyway still very very good.


We have seen some degradation with using a less performant device.
The Jitter test - which was meant to be very sensible to Jitter and other digital domain issues - shows some degradations, for the SPDIF converter (especially in the direction from Coaxial to TosLink).
Nevertheless, that doesn't really show in any of our classical analog measurements, which all score very good.
So it's probably benign.



3. Functionality

ADAT capabilities
Following a question of @Blumlein 88 , I'm happy to say that the small switch also works with 8 channels ADAT @48kHz

I've done a quick setup as follows:
PC generator > to 8 Dante channels (Dante controller) > to 8 ADAT output channels (RME 12Mic-D routing) > TosLink > TosLink switch > TosLink > UCX II

And we see the 8 channels in TotalMix

Port replicator capability of the SPDIF converter
The convertor always has both its outputs active.
When feeding the Coaxial/TosLink SPDIF converter with a TosLink input at 192kHz, the Coaxial output works perfectly, but the TosLink output's clock sometimes gets out of sync.
I didn't experience this issue at 44.1kHz, 48kHz and 96kHz though.
So it's acting perfectly as a converter, but it may give some issue to duplicate the output at high sampling frequency.


4. Conclusion

This review I thought would be quick and easy, and at the end, it took me almost the whole day...
But we learned a few interesting things.

After testing with the RME only, I wrote "Those devices are adding or substracting nothing to the signal they receive".

After further look and check with the more Jitter-sensitive (and way cheaper) Topping E30 II Lite, this is not absolutely true.
There are measurable performance degradations.

I still think we may trust those devices to do exactly what you want them to do.


So I would rate them as
Good for the job
(with this specific config)


Also, it's important to note that, even if those devices are not perfect, they absolutely don't prevent a well engineered device to perform perfectly through them.

So that's not -not only- those devices' "fault" if the Topping is giving some lower results with them.

Good engineering is not only about good raw figures.
It also gives you more freedom.

Your review came in handy: I had to figure out how to digitally connect my RME Fireface UCX II to my KEF LSX, which I just bought a week ago. The KEF uses Toslink SPDIF-in and I need the coaxial SPDIF-out of the RME Fireface UCX II to connect the KEF LSX to. So I got me the Toslink/Coaxial SPDIF converter.

 

[nagster]

Is it possible to measure delay on the cox-toslink conveter? Problems with lipsync on a TV for example.

I tried measuring the DUT delay of a relative of Rja4000's converter/switcher.

 

[Rja4000]

I tried measuring the DUT delay

5.2 micro second at 192kHz, that's one sample, isn't it ?

So just looping a Toslink cable on the AP and you add one sample delay.
And no additional delay at all while inserting any of the 2 devices.

While the RME, which is doing a real re-sampling, adds 3 or 4 samples.

Correct ?

All this confirms my measurements of the switch delay and what Matthias (@MC_RME ) wrote above.

Thank you very much for this useful addition.

 

[nagster]

View attachment 315795

5.2 micro second at 192kHz, that's one sample, isn't it ?

So just looping a Toslink cable on the AP and you add one sample delay.
And no additional delay at all while inserting any of the 2 devices.

While the RME, which is doing a real re-sampling, adds 3 or 4 samples.

Correct ?

All this confirms my measurements of the switch delay and what Matthias (@MC_RME ) wrote above.

Thank you very much for this useful addition.

thank you to you too.
Yes, I will post additional measurements later, but they seem to have little effect.
Regarding RME, I have attached part of the user's guide. By the way, RME's SRC was invalid in a series of measurements.

 

[nagster]

Recommended Optical Switch to use with ADI-2 DAC (Page 1) — ADI-2 & 2/4 Pro series, ADI-2 DAC series — RME User Forum

Jitter level sweep measurement.
I tried to get the measurement conditions closer to the linked one. I am always grateful to MC RME.

 

[e_h]

All TOSLINK POF (plastic optical fibre) indeed have a similar damping on the optical signal. And the longer the more damping, but 5 meter is still in normal condition and expected to work. Other materials like glass fibre, which is not the standard TOSLINK connection, have nearly no damping and can be used at much higher lengths, but can cause other problems due to internal different reflection and overdriving the receiver chip.

The issue with 192 kHz is typically the receiver. The transmitter, even if not specified for that data rate, often works ok.

@MC_RME please forgive my following stupid questions (and MS Paint schematics).
I'm considering buying the ADI-2 Pro FS R BE for use in the setup outlined in the attached jpeg ("SBX" refers to my SoundblasterX G6, used for virtual surround in headphones). PC source is used for music (mainly through speakers), movies and gaming (mostly headphones).

- What Toslink cable would you recommend, given the above quote?
- Will the device recommended by you (https://www.tindie.com/products/Beni_Skate/automatic-spdif-opticalrca-audio-switch/) work well in the below setup?
- You mentioned in the RME forum that "This unit has one single 'issue' for typical RME users: it doesn't speak ADAT, because it uses a special SPDIF transceiver inside." Is that relevant at all to my use case (if so, how)?
- Is the 192 kHz thing an issue at all in my case?

 

[MC_RME]

@MC_RME please forgive my following stupid questions (and MS Paint schematics).
I'm considering buying the ADI-2 Pro FS R BE for use in the setup outlined in the attached jpeg ("SBX" refers to my SoundblasterX G6, used for virtual surround in headphones). PC source is used for music (mainly through speakers), movies and gaming (mostly headphones).

- What Toslink cable would you recommend, given the above quote?

The cheapest one will do.

- Will the device recommended by you (https://www.tindie.com/products/Beni_Skate/automatic-spdif-opticalrca-audio-switch/) work well in the below setup?

Many people don't understand that the signal-dependent input switching does not relate to audio, but the SPDIF signal itself (carrier). So if your sources are all powered on there typically will be no switching. If only one is powered off/on the priority order of the switch can be used to make it work correctly.

- You mentioned in the RME forum that "This unit has one single 'issue' for typical RME users: it doesn't speak ADAT, because it uses a special SPDIF transceiver inside." Is that relevant at all to my use case (if so, how)?

Not rrelevant.

- Is the 192 kHz thing an issue at all in my case?

No.

 

[e_h]

Thank you for your replies, @MC_RME
I didn't quite understand this, so please bear with me.

Many people don't understand that the signal-dependent input switching does not relate to audio, but the SPDIF signal itself (carrier). So if your sources are all powered on there typically will be no switching. If only one is powered off/on the priority order of the switch can be used to make it work correctly.

- What do you mean by there will be no switching? Let's say I have both the Wiim and SBX connected to the Toslink switch in question, and they are both powered on, what will happen? I assumed that the device connected to the switch's "priority 1 port" would then get priority and put through its signal/audio through the switch to the RME, ignoring signals from the device connected to "priority 2 port" - is this not the case?
- If only one device is switched on I assume that device will get priority regardless of whether it's connected to the port with priority 1 or 2?

(The SBX device will always be on when the PC is on and feeding it current, which may very well be the case while I'm at the same time streaming music from the Wiim. There is no on/off button on the SBX and I don't want to physically unplug it from the PC every time I use the Wiim. From what I understand the Wiim has a standby mode where it doesn't output any signal. For this reason I figured I'd connect the Wiim to the switch's "priority 1 port", so that the Wiim always outputs audio to the speakers/headphones when it's on.)

 

[MCH]

Thank you for your replies, @MC_RME
I didn't quite understand this, so please bear with me.

- What do you mean by there will be no switching? Let's say I have both the Wiim and SBX connected to the Toslink switch in question, and they are both powered on, what will happen? I assumed that the device connected to the switch's "priority 1 port" would then get priority and put through its signal/audio through the switch to the RME, ignoring signals from the device connected to "priority 2 port" - is this not the case?
- If only one device is switched on I assume that device will get priority regardless of whether it's connected to the port with priority 1 or 2?

(The SBX device will always be on when the PC is on and feeding it current, which may very well be the case while I'm at the same time streaming music from the Wiim. There is no on/off button on the SBX and I don't want to physically unplug it from the PC every time I use the Wiim. From what I understand the Wiim has a standby mode where it doesn't output any signal. For this reason I figured I'd connect the Wiim to the switch's "priority 1 port", so that the Wiim always outputs audio to the speakers/headphones when it's on.)

Don't want to answer for @MC_RME , but what i understand, and based on my experience with other wm8805 based switches, not this one, is the following, in case it helps:
what the spdif transceiver reports is an input lock, and the way these switches work is to change the input and read if there is lock or not. But the transceiver does not have means to know if there is an actual audio signal playing or not. That means that if your source locks in, the switch will recognize it as a valid signal, no matter if music is playing or not. The problem is that some sources will lock even if nothing is playing, my tv locks even when switched off! (in standby mode). If you have one source that does this, as you say, you can put it to the lowest priority input, if you have more than one.... well, that's a bigger problem.
It is possible to actually read the i2s signal from the transceiver and search for a real audio input, but that is a different story and not possible with this particular device.

 

[Rja4000]

What do you mean by there will be no switching? Let's say I have both the Wiim and SBX connected to the Toslink switch in question, and they are both powered on, what will happen?

The switch you refer to is auto switching sources. There is no remote or select button.
The switch selects source based on which one is powered on and sends a stream through SPDIF or TosLink.
As @MCH explains above, this may be when it sends music or even in standby, depending of the source device.
I don't know (yet) what happens if 2 are sending stream. I guess it stays on the initial source until this one is stopping sending data.

So for your use, the switch I reviewed, with its remote and select button, is probably more adapted to your needs.
Of course, it's up to you to decide...