[boushh]

  I am Surrounded By Boxes od Multimode Fiber as we speak. OM3/ 50 Micron With LC Connectors. I work in the Storage Area Network realm. If you ditch the (IMHO Inferior for SAN) Ethernet Protocol,  SCSI 3 which is is very simple compared to the previously mentioned we are have Switches from Brocade which do 32GB/s per port. The OM3 Cables will work On the 4 Gb/s, 8 Gb/s 16 GB/s and 32 GB/s SFP's (Transceiver). Although what you see is that the Trasnmission Distance Decreases dramatically as you get up in speed.  
 
SCSI 3 is kind of like S/PDIF in that it works well for its intended purpose so it has not really changed the difference here as mentioned is the need for fast speeds in this field so The Transceivers and associated hardware has made vast improvements. I see Ethernet (TCP/IP) SAN's , baaaaaarffff as a complicated and lossy protocol better suited for the interwebs. the FC protocol using SCSI 3 does not tolerate losses, add un needed layers and is inherently simpler which means what you put in is what you get out and errr... Hmm maybe that's why I like Multibit.  I regress. Getting off topic.Where I was going was we have used the same cables for well over 10+ years but the optics and hardware is where we saw practically all the improvements. In about 10 years we went from 4 Gb/s Xfer rates being the norm to 32 Gb/s as the new standard. The cables have not changed much.  Maybe S/PDIF itself Does not need much tinkering, maybe the hardware needs an improvement.
 

 
Problem with SPDIF is the design, you can't beat physic
 
SPDIF had multiple Design goals:
 
1. Works on short length (big problem with real fibre, 50um or 62 um is not working under 2m (better 4m), the shorter the more problems with the "injected" light). POF with the big diameter hasn't that problem.
 
2. Should be easy to build and cheap
      - SPDIF connectors are not precise, an 50um or maybe 100um faser would not be aligned direct in front of the laser with such a connector (an 1 to 2 mm LED has of course no problem with these connectors)
      - SPDIF works with LED not with laser (cheaper, esier to design/build)
 
3. SPDIF should be robust in a home or studio environment
      - because of the big diameter, you can bend POF in a 2cm radius and it is still working (classical fibre bend radius > 10 or 15cm)
      - 1mm diameter means an dust partical on the fibre is no problem, again a classic optical fibre wouldn't work any more
 
3. SPDIF max length 50m (ok no problem with more recent technology here
 
4. Safety
      - SPDIF works in the visible range (650nm), classical fibres in the UV range (850/1300nm)
      - if one looks into the light source (you should never do this !!!), in the visible range the eye has a chance to react (close) before it is hurt, in the uv range, the eye could not react (becouse of that in medical systems, there is often a second light source in the visible range integrated into an laser/light emitter
)
 
POF (SPDIF's fibre) has a diameter of 1mm and an attenuation of around 150db/km, classical optical fibre has something around 3 (or lower) db/km (and 50 or 62 um diameter (core))
 
There were development builds with POF and 5GBit/s (100m) but than the POF gets complex and is'nt cheap any more (multi gradient POF, or multiple POF in a bundle for one direction).
 
SPDIF is really great for what it was designed for, but it is imho outdated and not useable for higher frequencys/bandwidth.
 
AT&T/ST is 62.5/125um fibre and the used connector is more precise, but the fibre has to be handled more carefully. It could be a great replacement, but I don't think there will be enough force in the industry to make a new standard  for optical digital connections. For most people HDMI,Electrical  connection or even USB is working well enough.
 
Small vendors will using better technology anyway (like candaeis, m2tech ... ), but that normally means it isn't compatible to other equipment.

 

[wink]

Quote:jonjen

 Ah, a fellow marketeer with a firm grasp of just what it takes to hit the mass market right where it lives… And don't forget those end of the fiscal quarterly specials and two-fers.
 
hahahahahahahahahaha 

 
That's it in a nutshell.    

 
You NEED to buy something nobody needs to get the deal sweeteners nobody wants.     

 
If you sell it, some nutcase will buy it - if you market it right.     

 
The three most important rules of sales - Hype, hype and more hype.    

 
I mean, what's more important than a set of steak knives...?   Answer: two sets of steak knives.    

 
Mine are special editions - they have fake ivory handles.  

 

 

[bigro]

   
Problem with SPDIF is the design, you can't beat physic
 
SPDIF had multiple Design goals:
 
1. Works on short length (big problem with real fibre, 50um or 62 um is not working under 2m (better 4m), the shorter the more problems with the "injected" light). POF with the big diameter hasn't that problem.
 
2. Should be easy to build and cheap
      - SPDIF connectors are not precise, an 50um or maybe 100um faser would not be aligned direct in front of the laser with such a connector (an 1 to 2 mm LED has of course no problem with these connectors)
      - SPDIF works with LED not with laser (cheaper, esier to design/build)
 
3. SPDIF should be robust in a home or studio environment
      - because of the big diameter, you can bend POF in a 2cm radius and it is still working (classical fibre bend radius > 10 or 15cm)
      - 1mm diameter means an dust partical on the fibre is no problem, again a classic optical fibre wouldn't work any more
 
3. SPDIF max length 50m (ok no problem with more recent technology here
 
4. Safety
      - SPDIF works in the visible range (650nm), classical fibres in the UV range (850/1300nm)
      - if one looks into the light source (you should never do this !!!), in the visible range the eye has a chance to react (close) before it is hurt, in the uv range, the eye could not react (becouse of that in medical systems, there is often a second light source in the visible range integrated into an laser/light emitter
)
 
POF (SPDIF's fibre) has a diameter of 1mm and an attenuation of around 150db/km, classical optical fibre has something around 3 (or lower) db/km (and 50 or 62 um diameter (core))
 
There were development builds with POF and 5GBit/s (100m) but than the POF gets complex and is'nt cheap any more (multi gradient POF, or multiple POF in a bundle for one direction).
 
SPDIF is really great for what it was designed for, but it is imho outdated and not useable for higher frequencys/bandwidth.
 
AT&T/ST is 62.5/125um fibre and the used connector is more precise, but the fibre has to be handled more carefully. It could be a great replacement, but I don't think there will be enough force in the industry to make a new standard  for optical digital connections. For most people HDMI,Electrical  connection or even USB is working well enough.
 
Small vendors will using better technology anyway (like candaeis, m2tech ... ), but that normally means it isn't compatible to other equipment.

​No Arguments here about the Optical S/PDIF its hardware or software. But There is Also Coax so the same protocol can be transported over different mediums. Copper Cables cable cannot handle 10Gb Ethernet over any distance most of what we see using 10GbE is Fiber now unless its localized runs. To the Systems its still using all the traditional Ethernet standards so no major changes there, the difference is the Transceivers in NICS or TOE's  and the transport medium.
 
For the Record I am not Copper Hating, Copper is still used with short run high speed interconnects Like Rapid IO and Infiniband. I don't see copper as inferior but within its limitations it does exceedingly well ( Rapid IO is 20Gb/s to (I believe now 100GB/s) Serial data with sub nanosecond latency over copper.  and I think it is Kick ass) Most of your post was on issues with the HW and Transport standard for the optical side of S/PDIF. and I Agree with you 100%,  there are far better ways to do optics.I don't claim to Be a S/PDIF or TCP or SCSI God for that matter and I am not saying Either is perfect but what I am saying is  S/PDIF the Protocol in of itself may not be the issue but the HW and Transport Medium Can Obviously be improved upon. On the Bandwidth, Usually the Limitations on Bandwidth are based on Hardware or transport medium not said protocol itself, I believe this to be the case with S/PDIF.

 

[boushh]

 
​No Arguments here about the Optical S/PDIF its hardware or software. But There is Also Coax so the same protocol can be transported over different mediums. Copper Cables cable cannot handle 10Gb Ethernet over any distance most of what we see using 10GbE is Fiber now unless its localized runs. To the Systems its still using all the traditional Ethernet standards so no major changes there, the difference is the Transceivers in NICS or TOE's  and the transport medium.
 
For the Record I am not Copper Hating, Copper is still used with short run high speed interconnects Like Rapid IO and Infiniband. I don't see copper as inferior but within its limitations it does exceedingly well ( Rapid IO is 20Gb/s to (I believe now 100GB/s) Serial data with sub nanosecond latency over copper.  and I think it is Kick ass) Most of your post was on issues with the HW and Transport standard for the optical side of S/PDIF. and I Agree with you 100%,  there are far better ways to do optics.I don't claim to Be a S/PDIF or TCP or SCSI God for that matter and I am not saying Either is perfect but what I am saying is  S/PDIF the Protocol in of itself may not be the issue but the HW and Transport Medium Can Obviously be improved upon. On the Bandwidth, Usually the Limitations on Bandwidth are based on Hardware or transport medium not said protocol itself, I believe this to be the case with S/PDIF.

 
I have no problem with the software side
 
I think it is (at the moment) limited to DTS/ATRAC/AAC (IEC61397) at the max, for newer formats the standard has to be improved. SPDIF (or AES/EBU) is (as long as it is not used for stereo 44/48 kHz). not more than a type of packet format, able to transport streams with multichannel data  (DTS/ATRAC...).
But I am afraid it has lost that fight against HDMI by now, I have never read of transporting the newer multi channel audio formats (DTS-HD, ATMOS ... all the stuff I no longer understand over SPDIF / AES/EBU. As long as HDMI is working, I am afraid there is no one working on a better SPDIF
 
Here are some informations about the used signal format:
http://www.hardwarebook.info/S/PDIF#Multi_channel_audio_and_S.2FPDIF
https://tech.ebu.ch/docs/tech/tech3250.pdf
 
(second one is aes/ebu but imho there is only a very limit difference in the format)

 

[bosiemoncrieff]

SF Opera bombed its first two productions of the season. Dream of the Red Chamber was a tedious retelling of the classical chinese novel in operatic form. It tried to summarize too much plot, and as a result, didn't pack the dramatic punch of a single dramatic action (ala Tosca, Tristan, Lear) until well into the second act. Andrea Chenier is classic wannabe Puccini. The production looked like it could have been at the 1896 premiere, and the direction had a vulgar tendency to play for the odd slapstick laugh—an old gardener wandering into a ballroom and taking off his shoes, to the horror of a fat female aristocrat, etc.
 
Don Pasquale, though, is excellent. Contemporary dress and intelligent direction make it a charming, tight comedy. No excess flab anywhere (thank you, Donizetti).

 

[bigro]

CentranceQuote:

   
I have no problem with the software side
 
I think it is (at the moment) limited to DTS/ATRAC/AAC (IEC61397) at the max, for newer formats the standard has to be improved. SPDIF (or AES/EBU) is (as long as it is not used for stereo 44/48 kHz). not more than a type of packet format, able to transport streams with multichannel data  (DTS/ATRAC...).
But I am afraid it has lost that fight against HDMI by now, I have never read of transporting the newer multi channel audio formats (DTS-HD, ATMOS ... all the stuff I no longer understand over SPDIF / AES/EBU. As long as HDMI is working, I am afraid there is no one working on a better SPDIF
 
Here are some informations about the used signal format:
http://www.hardwarebook.info/S/PDIF#Multi_channel_audio_and_S.2FPDIF
https://tech.ebu.ch/docs/tech/tech3250.pdf
 
(second one is aes/ebu but imho there is only a very limit difference in the format)

​Yes , if we are looking at newer formats there are obvious limitations. My understanding is that S/PDIF - AES/EBC will do 2 Channel uncompressed but has to compress all of the newer Multichannel formats. I am not to concerned with multi 5.x, 7.x.  multi channel world and like You I have don't understand too well, because its filled with buzzwords and by the time I gain an understanding of one there is some new format out.  HDMI is Convenient , It works, its good enough and its pay to play which makes the overlords of A/V happy. I do not do any serious listening with the super duper Multi Channel formats any way.
 
I would really like to see some new tech too. Pie in the Sky Alert.!!!! I would be really cool if some of the smaller audio companies Like Schiit, Emotiva, Salk, Centrance etc get together and work on an open type standard and interconnect that would improve upon the current standards that would really put a Kink in the pay for play armor. I know the Resources and costs are hard for smaller companies to swallow but they absolutely have the minds and the attitude to do it right and turn the audio world on its head.

 

[rtaylor76]

TOSlink is capable of 8 channels of 24-bit at 48kHz audio. This is ADAT lightpipe format. It's not an issue of bandwidth.

It is an issue of multi path and also fat st switching converting pulses to light, then receiving those pulses back to voltage pulses again. Two conversions that are unnecessary to me when you have coax available with the original voltage pulses from the source.

Don't want to lose that voltage you say? Input chips are usually pretty sensitive and don't even need that full 0.5V pulse...most can take down to 0.3V or even 0.2V and still work great with no drops in the bitstream.

Me personally, I run a -18db 75 ohm matched pad on my coax. This drops the noise floor by the same amount while keeping enough voltage pulse to work. Some have gone down to -24db pads with no issues, and some have found the limit at -20db. It depends on the output of your equipment, coax cable used, and input of the DAC. Of course, those with issues at lower db run many connectors and adapters (i.e. RCA to F, F pad, F to RCA). Just put the pad closest to the input. This really helps with computer systems and crazy grounds with USB systems. YMMV

I have also heard of taking the ground and one of the pins form AES and it will work on a S/PDIF input. I am not too sure of the specifications on AES, but I think it is a full 1V, so one leg makes sense. Not sure on this.

 

[boushh]

TOSlink is capable of 8 channels of 24-bit at 48kHz audio. This is ADAT lightpipe format. It's not an issue of bandwidth.

It is an issue of multi path and also fat st switching converting pulses to light, then receiving those pulses back to voltage pulses again. Two conversions that are unnecessary to me when you have coax available with the original voltage pulses from the source.

Don't want to lose that voltage you say? Input chips are usually pretty sensitive and don't even need that full 0.5V pulse...most can take down to 0.3V or even 0.2V and still work great with no drops in the bitstream.

Me personally, I run a -18db 75 ohm matched pad on my coax. This drops the noise floor by the same amount while keeping enough voltage pulse to work. Some have gone down to -24db pads with no issues, and some have found the limit at -20db. It depends on the output of your equipment, coax cable used, and input of the DAC. Of course, those with issues at lower db run many connectors and adapters (i.e. RCA to F, F pad, F to RCA). Just put the pad closest to the input. This really helps with computer systems and crazy grounds with USB systems. YMMV

I have also heard of taking the ground and one of the pins form AES and it will work on a S/PDIF input. I am not too sure of the specifications on AES, but I think it is a full 1V, so one leg makes sense. Not sure on this.

 
ADAT: 48kHz/24Bit*2 = 9.2 MBit/s (SPDIF is limited to around 5MBit/s), SACD/DVD-A Multichannel is max 9.8MBit/s, True-HD 19 MBit/s (i think Atmos is specified up to 48 MBit/s). So I have to disagree, Bandwidth is one of the problems here.
And ADAT (ligthpipe) is only specified up to 5m, may be enough for you, it is not enough for my scenario And you need higher quality fibres (POF or multicore fibre), than classical SPDIF.
(I know,  you can use repeaters but these mostly use better transportation methods (classic firbe ot CAT/5/6))
 
Switching between electical pulses and light beams may add run-time (but also is strengthen the signal for electrical transportation), but that is not such big a problem for a hifi system as it is for recording purposes (as long as the added run-time is stable). For simply playing it makes no difference if the signal arrives half a second later (as long as it is that time for all the signals/frequencies). The signal recovery has to be done for light and electrical pulses, so no difference here.
 
So yes you are partly right for audio up to 192kHz/24Bit you could build something with adat interfaces (not SPDIF, because different signals), but that would be the end (that is why MADI is using 50/125 fibre (when used optical)).

 

[spkalka]

Was a Manhattan concealed in the Schiit Schtack?

 

[jimmers]

  Was a Manhattan concealed in the Schiit Schtack?

What's after Manhattan, Berlin?

 

[bosiemoncrieff]

Odd that those are the only two places I want to move.

 

[jgreen16]

Odd that those are the only two places I want to move.

The only place I want to move to is Singapore! LOL.

 

[pdickerson]

The only place I want to move to is Singapore! LOL.

I second that. They know how to enforce the law.

 

[rtaylor76]

So yes you are partly right for audio up to 192kHz/24Bit you could build something with adat interfaces (not SPDIF, because different signals), but that would be the end (that is why MADI is using 50/125 fibre (when used optical)).

MADI rules!

As I also understand it, SACD/DVD-A theoretically has the bandwidth (barely) for 6 channels of high resolution audio, the bandwidth and data rate exceeded that of a DVD laser. This is why they use MLP (Meridian Lossless Packing). Either way, they turned off the digital outputs out of principal because the labels were too afraid of digital reproduction. It may of violated the Digital Millennium Copyright Act, manufacturers didn't care.

But back to the original point...

Glass SPDIF cables I hear can help. Not sure if they are worth the coin given that coax works just fine and much better option for way less money, at least for 2-channel resolution.

I wonder if DRM bits are still used by digital files anymore? I have a digital to digital converter that strips off the SCMS protection codes if I ever need it. I bought it for that reason, but it's main use is to convert SPDIF Coax, Toslink or AES/EBU to any of those other three formats.



Options for SCMS are: None, Pass-Thru, Original, and 1st Gen.

 

[artur9]

I wonder if DRM bits are still used by digital files anymore? I have a digital to digital converter that strips off the SCMS protection codes if I ever need it.

Some users (I think on the AES67 thread) have reported issues because the Yggy honors those bits and some of the pro equipment uses them. 
 
@Baldr
 
I wonder if bigger resistors work better than smaller ones in the R2R context?  I mean, sure, your DAC would be the size of a refrigerator but would it be more linear?