|
Post by Johnkenn on Feb 7, 2020 23:08:41 GMT -6
Well, it certainly sounds amazing now...I'll switch back at some point just to see.
|
|
|
Post by ragan on Feb 7, 2020 23:11:08 GMT -6
I totally get why they’re finicky and prone to breakage. I don’t get why a given cable could cause a stream of 1s and 0s to be altered in such a way as to become a different stream of 1s and 0s that sounds different. Spdif/AES are biphase encoded. This means that the bits are also the clock. The receiver will buffer the incoming signal, and lock a pll to the recovered clock to produce a secondary wordclock or system clock for the converters to utilize. Any missed bits means missed clock references and will allow the pll to start slightly "walking" away from the locked frequency. Do this often enough and you'll have a jittery system reference clock.. If you have a fractured cable you'll have internal reflections that might trigger the Schmitt inputs multiple times per edge and cause multiple error bits which will freak out the recovery plls. So if the pll becomes unreliable in the recovered signal we could start getting weird phase smear/time domain artifacts. Ok, that's something concrete. Re Schmitt triggering: you're talking specifically about a *damaged* optical cable right? One whose light pulses aren't strictly confined to the core anymore and are propagating unpredictably? This would be (I think) a totally different question than talking about allegedly "good" (ie expensive) optical cables vs allegedly "bad" (ie not expensive) optical cables as far as sonic performance. And for that matter, missing bits in general...this would also be related explicitly to cable (or at least connector) damage, right?
|
|
|
Post by BenjaminAshlin on Feb 8, 2020 3:23:12 GMT -6
Spdif/AES are biphase encoded. This means that the bits are also the clock. The receiver will buffer the incoming signal, and lock a pll to the recovered clock to produce a secondary wordclock or system clock for the converters to utilize. Any missed bits means missed clock references and will allow the pll to start slightly "walking" away from the locked frequency. Do this often enough and you'll have a jittery system reference clock.. If you have a fractured cable you'll have internal reflections that might trigger the Schmitt inputs multiple times per edge and cause multiple error bits which will freak out the recovery plls. So if the pll becomes unreliable in the recovered signal we could start getting weird phase smear/time domain artifacts. Ok, that's something concrete. Re Schmitt triggering: you're talking specifically about a *damaged* optical cable right? One whose light pulses aren't strictly confined to the core anymore and are propagating unpredictably? This would be (I think) a totally different question than talking about allegedly "good" (ie expensive) optical cables vs allegedly "bad" (ie not expensive) optical cables as far as sonic performance. And for that matter, missing bits in general...this would also be related explicitly to cable (or at least connector) damage, right? Like with most digital signals, you start with a sequence 0101000100111 and unless a cable is damaged or not up to spec the resulting output should be 0101000100111. A digital cable doesn't enhance the signal it can only degrade the signal. In the early 2000s we would have a lot of trouble with bad optical cables (protection&control systems) and quality ones were expensive. Now the cheap ones are much better, provided you don't put your oily fingers all over the connector.
|
|
|
Post by stormymondays on Feb 8, 2020 6:24:25 GMT -6
Oh please don’t make me crawl behind the desk to switch optical cables around!!!! 😅
|
|
|
Post by sozocaps on Feb 8, 2020 9:05:59 GMT -6
Needed a second optical cable, so I got a new one...just hooked it up...and I SWEAR things sound much wider. I'm not trying to promote that some cables sound better than others - I'm just wondering if maybe the cable I had had some issues or something. But I guess a digital cable doesn't have differing levels of qualities - they either work or not, right? Is there some kind've thing like "packet loss" they can suffer? (I have zero idea, so don't make fun of me lol) It probably just goes to show how fallible the human listening experience is...today I would swear it sounds better...probably just in my head. Guess I could switch back and forth...but that would take getting up lol. Maybe this has been your struggle all along... I remember never being happy with gear no matter how good it was supposed to be... and my friend is going through it now with his guitar rig. I just told him just because you use the same for example preamp does not mean you will have my sound. Everything else around the preamp is different. Cables ( I use Evidence Audio the best bar none IMHO), power amp, speakers, effects etc.... My sound is wide and clear his is a homogenized mess....
|
|
|
Post by Johnkenn on Feb 8, 2020 9:17:28 GMT -6
Needed a second optical cable, so I got a new one...just hooked it up...and I SWEAR things sound much wider. I'm not trying to promote that some cables sound better than others - I'm just wondering if maybe the cable I had had some issues or something. But I guess a digital cable doesn't have differing levels of qualities - they either work or not, right? Is there some kind've thing like "packet loss" they can suffer? (I have zero idea, so don't make fun of me lol) It probably just goes to show how fallible the human listening experience is...today I would swear it sounds better...probably just in my head. Guess I could switch back and forth...but that would take getting up lol. Maybe this has been your struggle all along... I remember never being happy with gear no matter how good it was supposed to be... and my friend is going through it now with his guitar rig. I just told him just because you use the same for example preamp does not mean you will have my sound. Everything else around the preamp is different. Cables ( I use Evidence Audio the best bar none IMHO), power amp, speakers, effects etc.... My sound is wide and clear his is a homogenized mess.... This is the first time I’ve used an optical cable to connect to a different DA...so not sure what my excuse has been in the past.
|
|
|
Post by sozocaps on Feb 8, 2020 9:41:06 GMT -6
Maybe this has been your struggle all along... I remember never being happy with gear no matter how good it was supposed to be... and my friend is going through it now with his guitar rig. I just told him just because you use the same for example preamp does not mean you will have my sound. Everything else around the preamp is different. Cables ( I use Evidence Audio the best bar none IMHO), power amp, speakers, effects etc.... My sound is wide and clear his is a homogenized mess.... This is the first time I’ve used an optical cable to connect to a different DA...so not sure what my excuse has been in the past. Ahahha but on topic yea optical cables are actually the worst for jiter as people have stated just because of the ADAT/TOSLINK format. The darn plug is not a positive connection and the old ADAT/TosLink fiber was not cut well like it is today. That being said I use some hot rod glass ones just to have the best I can for my apogee 16x RME interfacing... John
|
|
|
Post by Johnkenn on Feb 8, 2020 13:04:10 GMT -6
Oh please don’t make me crawl behind the desk to switch optical cables around!!!! 😅 Honestly, I just want to believe that it’s the cable and not just confirmation bias
|
|
|
Post by Johnkenn on Feb 8, 2020 13:05:07 GMT -6
And btw - the new cables I bought were from Amazon for $6
|
|
|
Post by stormymondays on Feb 8, 2020 13:08:18 GMT -6
Oh please don’t make me crawl behind the desk to switch optical cables around!!!! 😅 Honestly, I just want to believe that it’s the cable and not just confirmation bias If it’s an easy switch then you could actually test the cables I guess. Just send out and in some stuff and record it.
|
|
|
Post by sozocaps on Feb 8, 2020 16:16:49 GMT -6
Oh please don’t make me crawl behind the desk to switch optical cables around!!!! 😅 Honestly, I just want to believe that it’s the cable and not just confirmation bias Man a bad design reflective fiber optic with a horrible splice or termination polish in network land where I live can be horrible.... We use only corning just because of design and quality of the termination.
|
|
|
Post by Bob Olhsson on Feb 8, 2020 17:07:54 GMT -6
Old design SPDIF receivers generate massive quantities of clock jitter that most gear can't eliminate. I used to have fun moving optical cables around while people listened to the drifting soundstage. People forget that while bits are bits, the clock is still good ol' analog.
|
|
|
Post by Johnkenn on Feb 8, 2020 17:10:40 GMT -6
Btw - I’m clocking via word clock
|
|
|
Post by damoongo on Feb 9, 2020 1:29:35 GMT -6
Just send out and in some stuff and record it. This! Do it with each of the two cables in question. Then flip polarity on one of the recorded tracks during playback. Total null?
|
|
|
Post by Johnkenn on Feb 9, 2020 8:06:48 GMT -6
Just send out and in some stuff and record it. This! Do it with each of the two cables in question. Then flip polarity on one of the recorded tracks during playback. Total null? It’s a master DA to a pair of monitors. Guess I could monitor off the Apollo and route the DA back in to the Apollo...but that sounds like an awful lot of trouble.
|
|
|
Post by damoongo on Feb 9, 2020 16:03:30 GMT -6
This! Do it with each of the two cables in question. Then flip polarity on one of the recorded tracks during playback. Total null? It’s a master DA to a pair of monitors. Guess I could monitor off the Apollo and route the DA back in to the Apollo...but that sounds like an awful lot of trouble. Roger that. Some things are worth the trouble, and some aren’t! If you’re happy and don’t mind an unsolved mystery, then I say forge ahead! (But the nerd in me likes to definitively rule things out.)
|
|
|
Post by Martin John Butler on Feb 9, 2020 16:42:42 GMT -6
John, as they say, "don't worry, be happy".
|
|
|
Post by svart on Feb 10, 2020 8:49:19 GMT -6
When I designed my converters, I was told that I "needed to have optical or nobody will buy it". I refused, mainly because a lot of older SPDIF optical devices didn't work above 48KHz and the issues with poor transceivers as well as the poor interfacing issues like you're seeing here.
I got some grief from a few people for choosing to go COAX only, but I still stand by my decision. COAX will always be better than TOSLINK in any situation.
|
|
|
Post by Guitar on Feb 10, 2020 8:53:01 GMT -6
I think I heard Dan Lavry say that optical is the worst connection too, a long time ago. Apparently he has a 'white paper' that covers some of this.
The quote that I remember is something like, I'll paraphrase, "Any converter should be run on its own internal clock if possible. Any external clock can only be worse than the internal clock in terms of jitter, never better."
Then he gives some example of clocking over optical.
I hope I didn't butcher that.
|
|
|
Post by svart on Feb 10, 2020 9:11:31 GMT -6
I think I heard Dan Lavry say that optical is the worst connection too, a long time ago. Apparently he has a 'white paper' that covers some of this. The quote that I remember is something like, I'll paraphrase, "Any converter should be run on its own internal clock if possible. Any external clock can only be worse than the internal clock in terms of jitter, never better." Then he gives some example of clocking over optical. I hope I didn't butcher that. Nope, it's true. External wordclock can only ever be worse than internal clocks. However, there's a few different optical protocols, and some are very high quality. MADI is one, and it requires a very tight clock and high-quality optical cables. It uses true transmission rated fibers and transceivers. The MADI cable I use is only about 6ft long, dual fiber, cost about 50$, which was one of the cheapest ones I could find because these are "rated" for transmission unlike TOSLINK cables.
|
|
ericn
Temp
Balance Engineer
Posts: 14,928
|
Post by ericn on Feb 10, 2020 9:16:40 GMT -6
I think I heard Dan Lavry say that optical is the worst connection too, a long time ago. Apparently he has a 'white paper' that covers some of this. The quote that I remember is something like, I'll paraphrase, "Any converter should be run on its own internal clock if possible. Any external clock can only be worse than the internal clock in terms of jitter, never better." Then he gives some example of clocking over optical. I hope I didn't butcher that. Man I remember walking into a studio where everything was ADAT optical clocked over the fiber and simply running the clock over some of the cheapest RG 59 cables and you would have thought I re did the entire room. $30 worth of BNC video cables, needless to say the guy was a customer for life and still emails before buying anything!
|
|
ericn
Temp
Balance Engineer
Posts: 14,928
|
Post by ericn on Feb 10, 2020 9:25:30 GMT -6
I think I heard Dan Lavry say that optical is the worst connection too, a long time ago. Apparently he has a 'white paper' that covers some of this. The quote that I remember is something like, I'll paraphrase, "Any converter should be run on its own internal clock if possible. Any external clock can only be worse than the internal clock in terms of jitter, never better." Then he gives some example of clocking over optical. I hope I didn't butcher that. Nope, it's true. External wordclock can only ever be worse than internal clocks. However, there's a few different optical protocols, and some are very high quality. MADI is one, and it requires a very tight clock and high-quality optical cables. It uses true transmission rated fibers and transceivers. The MADI cable I use is only about 6ft long, dual fiber, cost about 50$, which was one of the cheapest ones I could find because these are "rated" for transmission unlike TOSLINK cables. We have to remember Toslink/ ADAT cables and interface were developed as a cheap consumer interface. Everybody at Sony always assumed if you cared you would go AES. The problem is besides being cheap ADAT is so damn compact, the fact that you can fit 16 ch of ADAT IO on a PCI card just makes it all to attractive. If either MADI or Dante would take hold in the low/ Mid priced recording world like they have in highend live and install the world would be such a better place!
|
|
|
Post by svart on Feb 10, 2020 9:38:06 GMT -6
Nope, it's true. External wordclock can only ever be worse than internal clocks. However, there's a few different optical protocols, and some are very high quality. MADI is one, and it requires a very tight clock and high-quality optical cables. It uses true transmission rated fibers and transceivers. The MADI cable I use is only about 6ft long, dual fiber, cost about 50$, which was one of the cheapest ones I could find because these are "rated" for transmission unlike TOSLINK cables. We have to remember Toslink/ ADAT cables and interface were developed as a cheap consumer interface. Everybody at Sony always assumed if you cared you would go AES. The problem is besides being cheap ADAT is so damn compact, the fact that you can fit 16 ch of ADAT IO on a PCI card just makes it all to attractive. If either MADI or Dante would take hold in the low/ Mid priced recording world like they have in highend live and install the world would be such a better place! The thing is, AES is spdif, but with the "professional" bit turned on.. there is no other real functional difference at the protocol level.
|
|
|
Post by Johnkenn on Feb 10, 2020 9:42:18 GMT -6
Since I’m clocking with word clock, I should be fine right?
|
|
|
Post by svart on Feb 10, 2020 9:48:26 GMT -6
Since I’m clocking with word clock, I should be fine right? Clocking a D/A with word and feeding it spdif at the same time? Technically not sure how that would work. Since spdif also contains it's own clock, you'd have to buffer the audio streams and realign them to the word clock.. it can be done, but you'd have some latency and chance for missing bits. Have you tried unhooking the word clock while playing through the converter to see if it's actually getting clocked by the word input? It should drop out if it is. If it doesn't, then it's likely selecting the spdif stream as the clock source automatically.
|
|