|
|
Post by Omicron9 on Jun 21, 2021 13:34:44 GMT -6
Greetings.
I've been reading the RGO threads regarding ADAT lightpipe. They are simultaneously fascinating and slightly confusing, but the more I read, the less confused I am. I think; not sure.
I need to link up a couple of things with lightpipe cables. Looking on amazon, I see plenty of options. The most obviously notable difference is metal connectors vs. plastic connectors. I know the metal isn't carrying signal, but wondering if one type of connector mates better with ADAT jacks. Or if it even matters.
Are there ADAT cables to avoid? Any to recommend?
With tiny red lights shining out of both ears,
-09
|
|
|
|
Post by tkaitkai on Jun 21, 2021 18:04:28 GMT -6
I use my Lynx Aurora as my A/D and go ADAT into an Apollo. I've tested a few different cables, and here's what I've found: - $10 Hosa cable: Sounds okay. Kind of hazy (god I hate that term), possibly because it's older and wasn't exactly treated with care. - $50 AudioQuest Forest cable: When I first tried this, I thought it sounded harsh and squeaky and unlistenable. I was also amazed that there was a difference at all (it's all just ones and zeros, right?). The more I tried it, the more I realized it was just more revealing in the top end, and was exposing some harshness that was being masked by the Hosa. - $100 AudioQuest Cinnamon cable: Bought this a few months ago to see if there was an improvement over the Forest cable. There was not. In fact, I couldn't detect any audible difference between the two. I returned it. I'm still curious to try some crazy, super high end audiophile cable, but at the same time, ADAT isn't exactly the most well-regarded protocol to begin with. At this point, I'm more interested in ditching ADAT altogether. Paging svart |
|
|
|
Post by notneeson on Jun 21, 2021 18:08:49 GMT -6
Paging Svart.
|
|
|
|
Post by jcoutu1 on Jun 21, 2021 18:18:03 GMT -6
|
|
|
|
Post by svart on Jun 21, 2021 18:28:45 GMT -6
Damn. I got multiple pages..
Anyway, the higher end cables are probably telling you what it really sounds like, but in a different way than most folks think.
The cheap cables tend to have bubbles, fractures, cracks, poorly mating connectors, etc. These all conspire to create reflections or "fuzzy" bits that might lead to missed or double bits in the stream.
That's a common problem, and I've run across it too. I had a cable that sounded funky and when I looked at the bitstream it was all messed up. Sometimes it would go unlocked for a few bits, which is not enough to glitch and crackle, but might be enough to sound strange.
Anyway, I cut the cable open and found bubbles and a few cracks from when I bent it a little too much.
Expensive optical cables don't make a properly working system work better. They're usually just better made than the cheap ones and don't break as easily IMHO.
|
|
ericn
Temp
Balance Engineer
Posts: 14,937 
|
Post by ericn on Jun 21, 2021 21:18:03 GMT -6
I have had to deal with 4 high end optical cables that just don’t fit right so be careful.
|
|
|
|
Post by superwack on Jun 21, 2021 23:31:56 GMT -6
I’ve seen debates on glass vs plastic toslink/ADAT cables - as is typical on the internet I have read that one is better (although which sometimes differs) and also that they are the same.
I’d appreciate Svart’s thoughts on it
|
|
ericn
Temp
Balance Engineer
Posts: 14,937
|
Post by ericn on Jun 22, 2021 5:55:05 GMT -6
I’ve seen debates on glass vs plastic toslink/ADAT cables - as is typical on the internet I have read that one is better (although which sometimes differs) and also that they are the same. I’d appreciate Svart’s thoughts on it Glass with glass transmitter and receiver Plastic with plastic transmitter and receiver is best. The problem is how do you know and what to do if they are mixed? Most ADAT light pipe boards use plastic because they are cheap. |
|
|
|
Post by svart on Jun 22, 2021 7:42:39 GMT -6
I’ve seen debates on glass vs plastic toslink/ADAT cables - as is typical on the internet I have read that one is better (although which sometimes differs) and also that they are the same. I’d appreciate Svart’s thoughts on it I've done some work with optical fibers in the past on high speed transmissions in the 2.5G range, so I have at least a little knowledge about fibers, so I'll talk about what I know. I don't see why a plastic fiber (acrylic/optical polycarbonate) would work any differently if there were no issues with inclusions (bubbles, cracks, etc) or the mating of the connectors and fiber end, at least at short distances. As I pointed out earlier, a fiber can't make a normally working connection sound better, it can only make it worse. Optical glass fiber is highly controlled during production and should be of higher quality which would result in fewer issues overall. I would think that the connectors would reflect the higher cost of glass fiber and be of higher quality too, but that's no guarantee because I'm sure there's no way to tell the overall quality unless you buy from a reputable brand. Glass has less attenuation over long distances than plastic, but we're talking hundreds to thousands of feet. ADAT/TOSLINK transmitters uses quite a bright LED, so for audio installation lengths, attenuation should be of zero concern. Some TOSLINK cables claim to have bundles of fibers for "low loss" which really doesn't mean anything here. The receivers generally have Schmitt triggers on their inputs so that any signal above a high threshold is a 1 and any signal below the low threshold is a 0. This removes ambiguity, noise and false triggers. The thresholds are usually fairly low, so there's plenty of headroom for an attenuated signal. Besides, if you're having issues with "low signal level" then your cable or connector is likely damaged. I would personally take a better single fiber cable over a cheaper multi-fiber cable. Care needs to be taken not to coil the cable too tightly, or bend or kink the cable more than a radius of maybe 3-4 INCHES. Tighter than that would surely break the cheaper plastic cables and possibly snap the glass fiber ones as well. I've seen folks coil up and bend TOSLINK cables like they were copper cables and wonder why they get iffy connections.. Glass fibers have polished ends that need to be cleaned every time you make a new connection. That maybe overkill for low speed audio, but for high speed data it's necessary. I've seen the really cheap ADAT cables have jagged ends where the clear plastic sticks out.. That's a huge indicator that the cable is trash. The cheap black cables you used to get with ADAT devices used to be much better quality than what you get today. I still have a few old ones and they still work OK. I bought a handful of newer cheap black cables and half of them either didn't fit the connectors or caused audio stream problems. They seem totally hit or miss. That being said, I think the best option is finding something in the 10-20$ range that has good reviews. At least at that price range you'd expect some QA and higher quality connectors to start. You could also buy a handful of the 1$ cables and just sort through them and find a few that work OK if you just wanted to. A few last notes.. I looked around on Amazon to see what's out there these days and sooo many of these cables have marketing horseshit like "gold coated connectors for best conduction" and "carbon fiber insulation for lowest noise" and other nonsense. A metal connector might make a better physical connection but there is absolutely no electronic reason for a metal shield/ground on an optical cable. As for insulation.. As long as it blocks light and keeps the cable from getting pinched or kinked, it's good. No need for fancy-pants carbon fiber weaving or whatever. I also didn't see many glass fiber cables and those that I did see were quite expensive. I guess TOSLINK is primarily plastic fiber these days, which is fine as long as it's quality is high enough that it doesn't ruin the physical mating or the digital bitstream. |
|
|
|
Post by svart on Jun 22, 2021 7:46:02 GMT -6
I’ve seen debates on glass vs plastic toslink/ADAT cables - as is typical on the internet I have read that one is better (although which sometimes differs) and also that they are the same. I’d appreciate Svart’s thoughts on it Glass with glass transmitter and receiver Plastic with plastic transmitter and receiver is best. The problem is how do you know and what to do if they are mixed? Most ADAT light pipe boards use plastic because they are cheap. I'm not sure there's any reason to believe that dissimilar materials would have an effect on each other. If a well polished plastic face met a glass face I wouldn't expect it to work any differently. But I do agree that almost all TOSLINK transceivers are going to be plastic connections and optics. |
|
rpc
Junior Member
Posts: 67
|
Post by rpc on Jun 22, 2021 8:24:44 GMT -6
Glass and plastic might have different index of refractivity, so you might get some reflections at the interface if your transceiver and fiber don't match. Likewise, the diameters of the cable and transceiver fiber should match or you'll get reflections. Ironically, due to total internal reflection the jacket should have no effect on the optical performance, so the primary criterion is physical protection of the fiber. A nice stiff cable that keeps you from bending it too tightly is actually best. As svart noted, the main thing is to get fiber that is free of irregularities (bubbles/cracks/density changes) and is cleanly cleaved/polished at the ends. Again ironically (I sense a theme here), those aren't the features touted by most audiophile cables.
RPC (who used to be in charge of server farms with far too many Fibre Channel hard drives)
|
|
ericn
Temp
Balance Engineer
Posts: 14,937
|
Post by ericn on Jun 22, 2021 8:41:51 GMT -6
Glass with glass transmitter and receiver Plastic with plastic transmitter and receiver is best. The problem is how do you know and what to do if they are mixed? Most ADAT light pipe boards use plastic because they are cheap. I'm not sure there's any reason to believe that dissimilar materials would have an effect on each other. If a well polished plastic face met a glass face I wouldn't expect it to work any differently. But I do agree that almost all TOSLINK transceivers are going to be plastic connections and optics. It’s what the guys who make them have told me for over 8 years & Im running the cheap ones they throw in with gear. |
|
|
|
Post by prene1 on Jun 22, 2021 19:47:46 GMT -6
I got mines from monoprice. Huge gauge momma’s but they go into my Cranbornes and apollos just fine.
|
|
|
|
Post by earlevel on Jun 28, 2021 15:03:59 GMT -6
I need to be a little contrarian here, feel free to tell me where I'm might be wrong. Possible issues with digital transmission over a cable are 1) data integrity (bits transmitting and being received correctly), 2) timing errors (jitter), and 3) electrical noise being transferred along with it, creeping into the analog section of the receiver. #3: Optical connectors give electrical isolation—this eliminates some of the issues that can come up with wired connections like USB. (I say "can" because a well-designed interface won't have issues with wired interfaces, it's a design flaw if they don't.) Isolation is inherent in optical cables, so this can't possibly be an issue with optical cables. #2: Light itself doesn't jitter, so the cable can't have different jitter than another. But it doesn't matter anyway, because interfaces buffer the data anyway—absolutely necessary because this is a packet-based protocol with multiple channels and auxiliary data. This can't possibly be an issue from one optical cable to the next. #1: So...when people talk about some optical connections having qualities of being less hazy, clearer sound, better sound stage, this implies a subtle change in the data. But when this is attributed to less than perfect optical conditions, why would this cause only subtle errors? Why not ear-drum-rattling pops when the MSB is interpreted by the receiver as a 1 instead of a 0? There is no reason, in the ADAT protocol, that only subtle data changes would occur. Let's look at #1 further, since it's the only one (so far) that's "extremely unlike", while the other two are "not possible": First, errors in the data have to be occurring if you're hearing something different. So, how often? If it's an error that happens once in several seconds, you're not going to hear it if it's one of the lowest bits, but higher bits would yield an impulse—it would be something like listening to pops and clicks on vinyl. If the errors happens many times a second, it would be possible to create a noise floor greater than that of the recording. But the catch here is that there is absolutely no reason that the errors would stay in the lowest bits, where the effect would be subtle. You would get, randomly, errors in high bits. You would recognize the cable as faulty, not just as having a "haze". This is just my intuitive appeal to why it's a practical impossibility. But, there is technical reason it can't happen, short of complete failure. ADAT protocol transport eight channels (for 48/44.1k) of 24-bit audio data in 256-bit packets. You get one 24-bit sample for each of the channels, per packet. The data is split into 4-bit nybbles, encoded NRZI, for by a synchronization bit of "1". The packet ends in some "0" synchronization bits, and a few user bits. Keep this in mind for a moment. Now, a lightpipe cable either passes light adequately or not. Yes, it's possible that one type (glass vs plastic), and length, could pass light more efficiently than another. But this is digital—the light is either bright enough to be read as a "1" or it's not. (Note: Because the data bits are sent NRZI, a "1" might be light or no light, it just depends on the order.) So, if a "bubble" or "dark spot" inhibits the light...it's going to do that for all "high" (light on) situations, including the sync bits. And it's likely to be absolutely constant (assuming you're not moving the cable). But in either case, synchronization is going to fail, because the sync bits won't get interpretted correctly (and if they don't cause the transceiver to block audio, all the bits will be in the wrong place anyway and utter chaos). The bottom line is this is digital, which either works or doesn't, AND it's isolated from analog noise (so don't confuse it with USB). It's going to either work, or not. (BTW, my buddies designed the protocol, I assure you they didn't overlook the integrity of the data stream.  ) PS—I realize that most of this discussion is on cable structural quality. My comment is aimed at the couple of references to sound quality. |
|
|
|
Post by svart on Jun 28, 2021 15:46:42 GMT -6
I need to be a little contrarian here, feel free to tell me where I'm might be wrong. Possible issues with digital transmission over a cable are 1) data integrity (bits transmitting and being received correctly), 2) timing errors (jitter), and 3) electrical noise being transferred along with it, creeping into the analog section of the receiver. #3: Optical connectors give electrical isolation—this eliminates some of the issues that can come up with wired connections like USB. (I say "can" because a well-designed interface won't have issues with wired interfaces, it's a design flaw if they don't.) Isolation is inherent in optical cables, so this can't possibly be an issue with optical cables. #2: Light itself doesn't jitter, so the cable can't have different jitter than another. But it doesn't matter anyway, because interfaces buffer the data anyway—absolutely necessary because this is a packet-based protocol with multiple channels and auxiliary data. This can't possibly be an issue from one optical cable to the next. #1: So...when people talk about some optical connections having qualities of being less hazy, clearer sound, better sound stage, this implies a subtle change in the data. But when this is attributed to less than perfect optical conditions, why would this cause only subtle errors? Why not ear-drum-rattling pops when the MSB is interpreted by the receiver as a 1 instead of a 0? There is no reason, in the ADAT protocol, that only subtle data changes would occur. Let's look at #1 further, since it's the only one (so far) that's "extremely unlike", while the other two are "not possible": First, errors in the data have to be occurring if you're hearing something different. So, how often? If it's an error that happens once in several seconds, you're not going to hear it if it's one of the lowest bits, but higher bits would yield an impulse—it would be something like listening to pops and clicks on vinyl. If the errors happens many times a second, it would be possible to create a noise floor greater than that of the recording. But the catch here is that there is absolutely no reason that the errors would stay in the lowest bits, where the effect would be subtle. You would get, randomly, errors in high bits. You would recognize the cable as faulty, not just as having a "haze". This is just my intuitive appeal to why it's a practical impossibility. But, there is technical reason it can't happen, short of complete failure. ADAT protocol transport eight channels (for 48/44.1k) of 24-bit audio data in 256-bit packets. You get one 24-bit sample for each of the channels, per packet. The data is split into 4-bit nybbles, encoded NRZI, for by a synchronization bit of "1". The packet ends in some "0" synchronization bits, and a few user bits. Keep this in mind for a moment. Now, a lightpipe cable either passes light adequately or not. Yes, it's possible that one type (glass vs plastic), and length, could pass light more efficiently than another. But this is digital—the light is either bright enough to be read as a "1" or it's not. (Note: Because the data bits are sent NRZI, a "1" might be light or no light, it just depends on the order.) So, if a "bubble" or "dark spot" inhibits the light...it's going to do that for all "high" (light on) situations, including the sync bits. And it's likely to be absolutely constant (assuming you're not moving the cable). But in either case, synchronization is going to fail, because the sync bits won't get interpretted correctly (and if they don't cause the transceiver to block audio, all the bits will be in the wrong place anyway and utter chaos). The bottom line is this is digital, which either works or doesn't, AND it's isolated from analog noise (so don't confuse it with USB). It's going to either work, or not. (BTW, my buddies designed the protocol, I assure you they didn't overlook the integrity of the data stream. ) PS—I realize that most of this discussion is on cable structural quality. My comment is aimed at the couple of references to sound quality. SPDIF and AES by extension are biphase encoded, which means that the data is also the clock signal. You're right, the receivers need a number of words to "lock" the PLLs to the timing once they've recognized a valid bitstream, but they generally do not buffer the data either. If you have reflections in the fiber, which ARE a perfectly valid and recognized source of false triggering, etc, they can "stumble" (for lack of a better technical term) the decoders so that the clocking gets out of sync with the data for a few bits before it relocks/recovers or before the PLL falls too far away in frequency to continue to decode data. It's this getting out of sync that can cause strangeness in the audio without going completely unlocked. Certain patterns of data can cause worse effects as well, which is why sometimes you get an unlock situation and a pop/click here and there. It's also why the pop/clicking tends to get worse before it goes completely unlocked as the system drifts further out of timing before going unlocked. Digital isn't always perfect. As I mentioned before, I've worked on high speed datacoms over fiber and it's quite common to get issues like the ones I've described despite it being "Ones and zeros" and theoretically impervious to misinterpretation. Anytime you move from one physical medium to another, there's always a chance for fuckery. |
|
|
|
Post by earlevel on Jun 28, 2021 15:47:42 GMT -6
Damn. I got multiple pages.. Anyway, the higher end cables are probably telling you what it really sounds like, but in a different way than most folks think. The cheap cables tend to have bubbles, fractures, cracks, poorly mating connectors, etc. These all conspire to create reflections or "fuzzy" bits that might lead to missed or double bits in the stream. That's a common problem, and I've run across it too. I had a cable that sounded funky and when I looked at the bitstream it was all messed up. Sometimes it would go unlocked for a few bits, which is not enough to glitch and crackle, but might be enough to sound strange. Anyway, I cut the cable open and found bubbles and a few cracks from when I bent it a little too much. Expensive optical cables don't make a properly working system work better. They're usually just better made than the cheap ones and don't break as easily IMHO. Good points, but I have to disagree with the one sentence about reflections, and missed or double bits. Light's too fast compared with the relatively slow bit rate to matter for reflections. (And light is going to reflect as light, so it wouldn't be a change anyway, without an incredible long cable.) And if you're going to get unlocked, it would just as likely be for the most significant bits as any others, so I'm not sure what you mean by "which is not enough to glitch and crackle". It would seem such errors would not likely be subtle, no? |
|
|
|
Post by earlevel on Jun 28, 2021 15:57:48 GMT -6
SPDIF and AES by extension are biphase encoded, which means that the data is also the clock signal. You're right, the receivers need a number of words to "lock" the PLLs to the timing once they've recognized a valid bitstream, but they generally do not buffer the data either. If you have reflections in the fiber, which ARE a perfectly valid and recognized source of false triggering, etc, they can "stumble" (for lack of a better technical term) the decoders so that the clocking gets out of sync with the data for a few bits before it relocks/recovers or before the PLL falls too far away in frequency to continue to decode data. It's this getting out of sync that can cause strangeness in the audio without going completely unlocked. Certain patterns of data can cause worse effects as well, which is why sometimes you get an unlock situation and a pop/click here and there. It's also why the pop/clicking tends to get worse before it goes completely unlocked as the system drifts further out of timing before going unlocked. Digital isn't always perfect. As I mentioned before, I've worked on high speed datacoms over fiber and it's quite common to get issues like the ones I've described despite it being "Ones and zeros" and theoretically impervious to misinterpretation. Anytime you move from one physical medium to another, there's always a chance for fuckery. OK, you addressed something I questioned in my reply above—you'd get pops and clicks—so we're in agreement there. I don't think we're too far off in what we're saying. I would agree with you on the reflections aspect if not for the fact ADAT is a relatively slow protocol (compared to the speed of light). You're not going to got a "light on" in a bit that should be dark, because of reflections, for instance. Again, I don't doubt that bad cables cause bad things. I'm mainly commenting to terms here like "hazy", "harsh", "revealing"–which we hear a lot for audio interconnects, and I believe apply to no ADAT cable (OK, possibly "harsh", for a very brief period before it gets thrown in the trash ). |
|
|
|
Post by cyrano on Jul 1, 2021 16:18:51 GMT -6
Glass is more fragile and expensive. No need for glass at ADAT speed and wavelength.
ADAT optical is a dinosaur in the fiber world. Not much to go wrong, really.
|
|
ericn
Temp
Balance Engineer
Posts: 14,937
|
Post by ericn on Jul 1, 2021 21:28:28 GMT -6
Glass is more fragile and expensive. No need for glass at ADAT speed and wavelength. ADAT optical is a dinosaur in the fiber world. Not much to go wrong, really. I don’t know clocking via ADAT always seams to screw the pooch ! |
|
|
|
Post by cyrano on Jul 2, 2021 14:13:26 GMT -6
20 years ago, clocks weren't as precise (and as cheap) as today. Much of ADAT's bad rep stems from these days and the Blackfaces.
I've got a bunch of ADAT gear, even old Fostex 16 bit thingies clock very well to RME's clock. I do on occasion come across gear that doesn't sync immediately. That's usually stuff that needs a new power brick, or a recap of the internal PSU. Or even a total recap.
That's the crux with early digital: you know it'll die. You just hope it doesn't die in an inconvenient moment.
|
|
|
|
Post by peterhess on Jul 2, 2021 17:12:55 GMT -6
I’ve been finding the occasional digital pop in tracks! Also, in simultaneous tracks (ie a mono source plus stereo room)... so it’s neither mic nor cable. From reading this, perhaps i have a compromised ADAT cable? ADAT from Apollo->mac.
Thinking about it, there’s ADAT from my 4-710->Apollo that’s seeing the stereo room. Mono straight to Apollo in. So if it’s on both those signals, it’s either the ADAT->mac, or something else entirely. I’ll clean and re-seat the connections.
|
|
)