|
|
Post by kcatthedog on Jan 25, 2024 13:29:53 GMT -6
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 25, 2024 14:56:33 GMT -6
Nice it now has the better Aurora (N) clock! I presume this was coming eventually.
but ewww they let the user select minimum phase filters to phase and alias their recording. Gross. Stuff like this and oversampling filters should be out of the hands of the end user. I've seen people mess up the synchrolock on Lynx cards and Hilo too so their stereo imaging is poor.
I like the Apogee and Crane Song approaches of not letting the end user or audiophile tweaker change anything. Just have it work. They don't even have a functional master clock selection like the Dangerous Converts do because they're asynchronous.
Dan
|
|
|
|
Post by ml on Jan 25, 2024 15:35:23 GMT -6
|
|
|
|
Post by robo on Jan 25, 2024 15:36:41 GMT -6
Interesting!
I can’t say I have any complaints about the sonics of my V1, but I’m glad they’re committed to the format.
I wish they had included more expandability. 16 channels isn’t enough for me, and probably lots of others. I have to run multiple interfaces.
|
|
|
|
Post by Johnkenn on Jan 25, 2024 18:22:32 GMT -6
Nice it now has the better Aurora (N) clock! I presume this was coming eventually. but ewww they let the user select minimum phase filters to phase and alias their recording. Gross. Stuff like this and oversampling filters should be out of the hands of the end user. I've seen people mess up the synchrolock on Lynx cards and Hilo too so their stereo imaging is poor. I like the Apogee and Crane Song approaches of not letting the end user or audiophile tweaker change anything. Just have it work. They don't even have a functional master clock selection like the Dangerous Converts do because they're asynchronous. Dan Dude...do you shit on everything? |
|
|
|
Post by the other mark williams on Jan 25, 2024 18:34:14 GMT -6
Nice it now has the better Aurora (N) clock! I presume this was coming eventually. but ewww they let the user select minimum phase filters to phase and alias their recording. Gross. Stuff like this and oversampling filters should be out of the hands of the end user. I've seen people mess up the synchrolock on Lynx cards and Hilo too so their stereo imaging is poor. I like the Apogee and Crane Song approaches of not letting the end user or audiophile tweaker change anything. Just have it work. They don't even have a functional master clock selection like the Dangerous Converts do because they're asynchronous. Dan I think giving the user a choice is really cool. I believe that’s the eventual plan for the Metric Halo MkIV interfaces. The minimum phase option gives less latency, which is really welcome in the composition phase. But then if one has the option of switching to linear phase during the mix (if one likes that sound better), I think that’s great. |
|
|
|
Post by audiospecific on Jan 25, 2024 19:00:57 GMT -6
I was impressed with the first version of them. Help someone set up one with Dante Rednet I/o. It has very good mixer inside that can be used for monitoring. Been contemplating the Hilo with a dante card vs. the DAD Core256 for my studio build. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 25, 2024 20:03:11 GMT -6
Nice it now has the better Aurora (N) clock! I presume this was coming eventually. but ewww they let the user select minimum phase filters to phase and alias their recording. Gross. Stuff like this and oversampling filters should be out of the hands of the end user. I've seen people mess up the synchrolock on Lynx cards and Hilo too so their stereo imaging is poor. I like the Apogee and Crane Song approaches of not letting the end user or audiophile tweaker change anything. Just have it work. They don't even have a functional master clock selection like the Dangerous Converts do because they're asynchronous. Dan Dude...do you shit on everything? Lynx is great but it’s weird to present the user with a choice here. Some things are better not left to the end user. |
|
|
|
Post by sirthought on Jan 25, 2024 23:31:51 GMT -6
Dude...do you shit on everything? Lynx is great but it’s weird to present the user with a choice here. Some things are better not left to the end user. That didn't answer his question. Do you shit on everything? |
|
|
|
Post by suicity on Jan 26, 2024 3:23:03 GMT -6
Dan is the RGO's Resident Cantankerous Old Man
|
|
|
|
Post by audiospecific on Jan 26, 2024 11:10:58 GMT -6
Dan is the RGO's Resident Cantankerous Old Man I'm thinking the Simon Cowell of audio. |
|
|
|
Post by tkaitkai on Jan 26, 2024 11:18:36 GMT -6
Hopefully this means I can get a V1 for $500 now.  |
|
|
|
Post by nudwig on Jan 26, 2024 11:37:43 GMT -6
Lynx is great but it’s weird to present the user with a choice here. Some things are better not left to the end user. That didn't answer his question. Do you shit on everything? TDR is totally Dan recommended (not a dig as it got me back using them regularly, just helping chessparov out as he must be busy to miss this low hanging jokefruit) |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 26, 2024 13:46:34 GMT -6
Lynx is great but it’s weird to present the user with a choice here. Some things are better not left to the end user. That didn't answer his question. Do you shit on everything? No. There's a lot of good sounding and useful pieces and some superb stuff but we've come to the point where letting people tweak certain things is insane. Good luck to your average user with an ever-decreasing as the years march on skill and knowledge level to get the same results with Izotope RX's SRC with it's crazy customizability that could be obtained with a click or two from a SoX based or r8brain src just tweaking the filters that should be set for them by the manufacturer or programmer. Minimum phase filters with thousands of degrees of phase shift in the audible bandwidth have no reason to be in anything high-fidelity.
Now we even have cutting edge tools meant to combat the ineffective band rejection of poorly chosen filters, the horrible marketing machine of "high-resolution audio", the non-linear pumping of abused well-marketed but antiquated compressors, and to broadcast material mastered ahem butchered by the popular music industry. We can purchase good tools to combat the bad tools (and good tools with a much more limited range of useful operation) abused by others. That is how insane the modern audio gear market is.
|
|
|
|
Post by the other mark williams on Jan 26, 2024 20:41:18 GMT -6
I gave a reasonable explanation for why someone might choose to use minimum phase mode over linear phase mode.
I’ll mention a different example that is contrary to Dan’s tirade.
Metric Halo had to switch from AKM converters (MkIII interfaces) after the AKM fire. They decided on TI chips for various reasons. The new TI chips can be run in either linear phase mode (which is what the AKM chips in MkIII interfaces had) or minimum phase mode. MH did extensive listening tests and everyone at the company thought the minimum phase mode in the TI chips sounded better. So that’s what they went with. After the fact, some people on the MIO Listserv asked if MH would consider giving the end user the choice to use the linear phase mode if they preferred it. MH added it as a feature request and are considering the best way to implement the feature.
From what I can tell, Dan is saying that manufacturers should not give end users a choice. But that if those same manufacturers happen to disagree with him on which mode sounds better, the manufacturers must be tin-eared dumbasses.
I don’t know what else to say here.
|
|
|
|
Post by Hudsonic on Jan 27, 2024 7:07:15 GMT -6
It's a pretty bad idea to offer choice of filters in dacs certainly. The reason is that the dac is supposed to reproduce what it is fed. Changing the filters gets away from proper design that should be set to allow the box to reproduce what it is fed. A dac is not a color box. All the things before the dac are potential color / tone boxes.
A so-called "feature" that is bound to confuse and confound.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 27, 2024 13:16:00 GMT -6
It's a pretty bad idea to offer choice of filters in dacs certainly. The reason is that the dac is supposed to reproduce what it is fed. Changing the filters gets away from proper design that should be set to allow the box to reproduce what it is fed. A dac is not a color box. All the things before the dac are potential color / tone boxes. A so-called "feature" that is bound to confuse and confound. Yes and having a thousand degrees or more of phase shift is not what you want in an anti-alias filter. Even if the designer chooses a fir filter with inadequate on paper band rejection, this can still reject almost all the frequencies near nyquist in real recordings where the mics and older electronics roll off while having less ringing or processing time than steeper fir filters. That's the Weiss approach. Of course sometimes the steeper fir filter is necessary for digital vomit, synths, rejecting harmonics of processors, etc. And of course double sample rates relax the filter requirements but while 50-60 khz might be better for the FIR filter as Daniels Lavry and Weiss found out in the 90s and everybody agreed with, that can present playback systems with issues. Still you can't have an analyzer in-between the converter and whatever it's feeding so having selectable filters or god forbid, an Izotope RX SRC or TDR Ultrasonic type interface, is just overkill. There are no tonal tweaks to be had because it's a technical issue that should be solved for the user by the equipment designer or coder.
Take the Genelec the ones. The custom coax mid-tweeter assembly is super cool but has headroom issues already in the larger monitors meant to be used as mains with the sub stack and it breaks up hard at about 28 kHz unlike your ears. Genelec runs their internal DSP for it at 96 kHz and the woofer at 48 so if you feed from a higher sample rate than 48 kHz to try bypass the asynchronous sample rate conversion (you can't) and avoid the anti-imaging filter being applied to get rid of the mirror images when upsampled, you can inadvertently be feeding it content you cannot hear that will intermodulate with the hardbreak up back into the audible band you can hear. A soft dome just will run out of headroom and be in more danger of cooking. Often this is present in non-linear processors (even if anti-aliased or analog) and even some older recording technology could capture some high frequencies that would get boosted by shelves and distorted by various processors. The guy working at 44.1 or 48 kHz or with some garbage old lofi opamp console won't be distorting his monitoring as much as the 88.2 / 96 kHz guy with the good computer and clean circuitry.
|
|
|
|
Post by ab101 on Jan 27, 2024 20:15:17 GMT -6
The original Lynx Hilo was supposed to be slightly better audiowise than the Aurora (n). So, if this Lynx Hilo 2 is better than the Lynx Hilo original, then that is saying something.
As to the phase modes, they do not have to be used, so that should solve that problem, unless I am misunderstanding it.
It will be interesting to hear comparisons of this with other top AD/DA units.
|
|
ericn
Temp
Balance Engineer
Posts: 15,011 
Member is Online
|
Post by ericn on Jan 27, 2024 22:03:17 GMT -6
It's a pretty bad idea to offer choice of filters in dacs certainly. The reason is that the dac is supposed to reproduce what it is fed. Changing the filters gets away from proper design that should be set to allow the box to reproduce what it is fed. A dac is not a color box. All the things before the dac are potential color / tone boxes. A so-called "feature" that is bound to confuse and confound. I predict weekly threads on various forums of which mode is better. I agree the designers should have made a choice, it should be part of the design. The biggest problem here is maybe 10% of the users will have any idea of what they are doing. Options always seam attractive, till you realize it’s a time sync and worse most of the guys selling it are going to have no idea what it all means except for what ever blurb Lynx provides. If this were in the world of contracting/ install this would be so deep in the utilities menu most would never find it. |
|
|
|
Post by aremos on Jan 28, 2024 0:03:08 GMT -6
Love the original Hilo & has become the "center piece" (main converter & monitor controller) of my studio.
|
|
|
|
Post by robo on Jan 28, 2024 0:49:57 GMT -6
I don’t know enough about filters in DAC design to have an opinion, but Lynx doesn’t strike me as a company to include the feature if there wasn’t a compelling reason to do so. It’s not the sort of thing high-end interface users are clamoring for, after all.
Remember there are two separate DA converters in the Hilo. One for monitoring and one for analog line out. Perhaps there’s something subjectively nice about the minimum phase option for the analog loop? It wouldn’t matter if it was as accurate in that context.
|
|
|
|
Post by ab101 on Jan 28, 2024 11:09:06 GMT -6
It's a pretty bad idea to offer choice of filters in dacs certainly. The reason is that the dac is supposed to reproduce what it is fed. Changing the filters gets away from proper design that should be set to allow the box to reproduce what it is fed. A dac is not a color box. All the things before the dac are potential color / tone boxes. A so-called "feature" that is bound to confuse and confound. I predict weekly threads on various forums of which mode is better. I agree the designers should have made a choice, it should be part of the design. The biggest problem here is maybe 10% of the users will have any idea of what they are doing. Options always seam attractive, till you realize it’s a time sync and worse most of the guys selling it are going to have no idea what it all means except for what ever blurb Lynx provides. If this were in the world of contracting/ install this would be so deep in the utilities menu most would never find it. I agree! Meanwhile, the Lynx support people are fantastic and when I talk to them, they are more than willing to help me with achieving the best quality. I imagine they will have an FAQ or maybe something in the PDF manual that will explain what to do. |
|
|
|
Post by kcatthedog on Jan 28, 2024 11:27:49 GMT -6
Agreed, excellent support here too with my N, pretty cool to know that principles in the company are doing support!
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 28, 2024 12:21:44 GMT -6
I don’t know enough about filters in DAC design to have an opinion, but Lynx doesn’t strike me as a company to include the feature if there wasn’t a compelling reason to do so. It’s not the sort of thing high-end interface users are clamoring for, after all. Remember there are two separate DA converters in the Hilo. One for monitoring and one for analog line out. Perhaps there’s something subjectively nice about the minimum phase option for the analog loop? It wouldn’t matter if it was as accurate in that context. I mean the concept of a filter is relatively straight forward, either remove or get the crap caused by conversion like aliasing outside the realms of human hearing. Unfortunately filters affect both amplitude and phase, plus they introduce their own issues like passband ripple (rotating / oscillating gain from unity) and in terms of minimum phase potential group delay's at different bandwidths, so you have inconsistent phase but less latency.
Now, a linear phase filter will delay certain bandwidths (frequencies) by phase shifting until everything is back in order but of course that adds additional delay. When it comes to latency one has to factor in everything though, despite filters adding delay's transmission methodologies, drivers etc. account for most of it and it's like developing a sports car where you shave 20 grams off the mirror then replace the boot floor with cardboard. Also this really only affects recording as in most cases playback with 20ms of latency just isn't an issue, so, usually I don't see a good enough reason to not use a half band or steep (sharp) linear.
Back in the day's where we hunter gathered and pentium's with turbo boosts existed min phase made more sense. Nowaday's just record at 96Khz, even with an inbuilt soundcard via core audio it yields a 5.9ms RTT @ 64 samples or even better use DSP mixers for cue's. Sure, there has been some testing that says even a 3ms latency can affect a singer but we all have own opinions on that one. Ultimately, I get where @tomegatherion is coming from and it's not a complaint because you just stick to the manufacturer default. Ultimately though even if I do (sort of) understand why you'd allow a selectable filter the issue is, if you have the technical knowledge to even at a fringe level understand this stuff than what about Thunderbolt, DSP mixers etc.? Why would you select a min phase filter to try and reduce group delay when there's better options? I might be missing something obvious here but it doesn't entirely make sense.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 28, 2024 12:42:45 GMT -6
Oh and the reason I'm getting into the technical stuff is because most of us know by now Lynx or Metric Halo are no brainers really. They'll last forever, sound great, support is top notch and if you can afford one / want one then you buy it.
|
|
