ericn
Temp
Balance Engineer
Posts: 16,098
|
Post by ericn on Jan 28, 2024 12:43:50 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.
An educated guess as to why. The Hilo has achieved a following in Audiophile circles, these same circles have a long, long debate about what type of filter is best. A single device that can serve both sides of the debate and can contribute to the debate because other than the filter type everything else is static, it will sell.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 28, 2024 15:22:04 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.
the only reason we have digital audio tools that are just as useful as most analog or surpass most analog tools is FIR filters. For someone to make the preposterous claim that the they’re bad and to condemn them suddenly for the final anti-alias filter of their converter is fucking nuts especially when the signal has already passed through numerous fir filters: what is controlling the multiplier of the signal (lookahead dynamics, multirate side chains, all modern digital limiters), correcting the iir filters (running them at a higher sampling rate or in the handleful of digital eqs that use a fir filter to correct the cramped iir filters at single sample rates. The latter can sound pretty good but aren’t as accurate as just running the iir filter at a higher sample rate and it might be automated), or creating the pcm files that are being manipulated in the daw and played back by the da converter. There are some plugin developers that use iir anti-alias filters in their algorithms but they, eg u-he, overloud, softube, mostly target their products at vsti musicians and started in a time of much less powerful CPUs when running an fir filter at a low buffer at 88.2 or 96 khz would grind a computer to a halt. Their plugins for musicians still sound pretty good but only their plugins that do not shift the phase (or are part of phase scrambling fx, eg reverbs, delays, old broadcast processor emulation, tape) work easily in a mix on recordings of real instruments without screwing things up.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 28, 2024 15:49:59 GMT -6
For someone to make the preposterous claim that the they’re bad and to condemn them suddenly for the final anti-alias filter of their converter is fucking nuts especially when the signal has already passed through numerous fir filters: Who's saying that exactly?
|
|
|
Post by viciousbliss on Jan 28, 2024 16:14:30 GMT -6
People on Head-fi buy Hilos. They love to mess with stuff. A lot of people on audiophile forums are high on compulsion and low on objective audio knowledge. They are always toying with things, buying thousands of dollars of gear, etc, because they're trying to solve problems caused by flaws in the designs, impedance mismatches, and things like that. I've always appreciated the candor of guys like Dan because these companies need to be held accountable. These are business relationships we have with them and we can't lose sight of that.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 28, 2024 16:20:00 GMT -6
Well, it seems like outside of pro audio there's a market I don't get. I'd just stick to some form of linear phase when it comes to converters, it's done wonders for many years and if latency is an issue (in general) then use a DSP mixer or up the sample rate. There's really no issue to solve here in my book..
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 28, 2024 21:21:26 GMT -6
For someone to make the preposterous claim that the they’re bad and to condemn them suddenly for the final anti-alias filter of their converter is fucking nuts especially when the signal has already passed through numerous fir filters: Who's saying that exactly? People who believe that maybe cutting out the inaudible content in a way that is certain to distort the audible content improves the fidelity of the audible content.
|
|
|
Post by notneeson on Jan 28, 2024 21:39:53 GMT -6
I dunno, I find it hard to believe that the Hilo comes with a built in “suck” button. It may be suboptimal, in theory, but so is external clocking.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 28, 2024 21:46:35 GMT -6
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.
An educated guess as to why. The Hilo has achieved a following in Audiophile circles, these same circles have a long, long debate about what type of filter is best. A single device that can serve both sides of the debate and can contribute to the debate because other than the filter type everything else is static, it will sell. Lynx also get to claim lower round trip latency over usb for the vsti musicians with computers too crappy to run at low buffer at double sampling rates.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 28, 2024 22:47:15 GMT -6
I dunno, I find it hard to believe that the Hilo comes with a built in “suck” button. It may be suboptimal, in theory, but so is external clocking. Lynx letting the user turn Synchrolock off has led to over a decade of misguided forum posts and flame wars...
|
|
|
Post by notneeson on Jan 28, 2024 23:40:30 GMT -6
I dunno, I find it hard to believe that the Hilo comes with a built in “suck” button. It may be suboptimal, in theory, but so is external clocking. Lynx letting the user turn Synchrolock off has led to over a decade of misguided forum posts and flame wars... Happily not on this rather level headed forum.
|
|
|
Post by Johnkenn on Jan 29, 2024 16:38:35 GMT -6
I dunno, I find it hard to believe that the Hilo comes with a built in “suck” button. It may be suboptimal, in theory, but so is external clocking. Lynx letting the user turn Synchrolock off has led to over a decade of misguided forum posts and flame wars... I’m not arguing any of this - it’s above my head. But your responses in the last month have been - well…kind of arrogant. Not sure what’s going on, but a little humility would go a long way.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 29, 2024 20:18:01 GMT -6
I dunno, I find it hard to believe that the Hilo comes with a built in “suck” button. It may be suboptimal, in theory, but so is external clocking. It's not a "suck" button though, various methodologies have different trade off's. More speed for less accuracy, the dork in me relates it to building a new character for an RPG game where you might have a lot of strength but the trade off is agility. You only get a certain number of "points" to use and there is no free lunch. Whether someone cares, notices, enjoys it more I honestly could never say or ever attempt to guess. Although I will say nothing about this is theoretical, filters wouldn't exist in the first place otherwise, they're created to solve a technical issue and there's multiple ways to go about it. External clocking is also not theoretical, at best it doesn't change anything and at worst it can mess up your clocking. What we seem to struggle with is separating technicalities from preference, converters are as a simile like doctors in which their purpose is to do no harm. It doesn't mean that after your appointment you can't pickup a pack of 20, go to the nearest bar and order 5 shots. Some might prefer that, some definitely won't but if it's the converter mangling things you're stuck with that. Take another example, standalone I prefer the sound of my MOTU 1248 to the Carbon and I know the former is technically inferior. However when I mangle the sound with the SSL the Carbon works best because of the level of transparency whereas the MOTU imparts a sound of its own as well. Technically superior doesn't often equate to "better" sounding but all these super smart people who create this stuff do it for a reason.
|
|
|
Post by notneeson on Jan 29, 2024 21:44:17 GMT -6
I dunno, I find it hard to believe that the Hilo comes with a built in “suck” button. It may be suboptimal, in theory, but so is external clocking. It's not a "suck" button though, various methodologies have different trade off's. More speed for less accuracy, the dork in me relates it to building a new character for an RPG game where you might have a lot of strength but the trade off is agility. You only get a certain number of "points" to use and there is no free lunch. Whether someone cares, notices, enjoys it more I honestly could never say or ever attempt to guess. Although I will say nothing about this is theoretical, filters wouldn't exist in the first place otherwise, they're created to solve a technical issue and there's multiple ways to go about it. External clocking is also not theoretical, at best it doesn't change anything and at worst it can mess up your clocking. What we seem to struggle with is separating technicalities from preference, converters are as a simile like doctors in which their purpose is to do no harm. It doesn't mean that after your appointment you can't pickup a pack of 20, go to the nearest bar and order 5 shots. Some might prefer that, some definitely won't but if it's the converter mangling things you're stuck with that. Take another example, standalone I prefer the sound of my MOTU 1248 to the Carbon and I know the former is technically inferior. However when I mangle the sound with the SSL the Carbon works best because of the level of transparency whereas the MOTU imparts a sound of its own as well. Technically superior doesn't often equate to "better" sounding but all these super smart people who create this stuff do it for a reason.
We are saying basically the same things. I'm using "theoretical" purely in the sense of "should sound WAY better in theory" but in practice preferences vary. I am not here to talk about Nyquist... ever.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 29, 2024 22:20:33 GMT -6
We are saying basically the same things. I'm using "theoretical" purely in the sense of "should sound WAY better in theory" but in practice preferences vary. I am not here to talk about Nyquist... ever. Cool and by the way, it's hard to relay in text but I don't take any of this seriously. I have a passing technical interest in audio due to previous jobs but I don't even do that any more. I could just as happily chat about car engines or TV shows over a cold one..
On a public forum even if it's a direct quote I'm aware that others read it and hopefully this topic helps someone?! It's nice to bridge the realms of technical vs. reality because many, many years ago I found it confusing personally. I can't say having the answers makes me a better mixer though .. Ultimately, it's something to chat about and if Nyquist ain't your thing okay, if it's someone else's I'll happily talk to them about it.
|
|
|
Post by notneeson on Jan 29, 2024 22:44:26 GMT -6
We are saying basically the same things. I'm using "theoretical" purely in the sense of "should sound WAY better in theory" but in practice preferences vary. I am not here to talk about Nyquist... ever. Cool and by the way, it's hard to relay in text but I don't take any of this seriously. I have a passing technical interest in audio due to previous jobs but I don't even do that any more. I could just as happily chat about car engines or TV shows over a cold one..
On a public forum even if it's a direct quote I'm aware that others read it and hopefully this topic helps someone?! It's nice to bridge the realms of technical vs. reality because many, many years ago I found it confusing personally. I can't say having the answers makes me a better mixer though .. Ultimately, it's something to chat about and if Nyquist ain't your thing okay, if it's someone else's I'll happily talk to them about it.
All good, my dude. What I meant is that I’m interested in the more practical side or making music and recordings. Probably should have added a smiley face or something.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 31, 2024 8:54:10 GMT -6
Lynx letting the user turn Synchrolock off has led to over a decade of misguided forum posts and flame wars... I’m not arguing any of this - it’s above my head. But your responses in the last month have been - well…kind of arrogant. Not sure what’s going on, but a little humility would go a long way. been a rough few months and I’m disgusted with a lot of shit: gear, gear marketing, ersatzes, artists not ready to play live yet alone to go to the studio but they went to the studio with the music not even being written, certain manufacturers, labels, and promoters. Sorry.
|
|
kcatthedog
Temp
Super Helpful Dude
Posts: 16,049
Member is Online
|
Post by kcatthedog on Jan 31, 2024 9:05:52 GMT -6
Hey Dan, sounds like a ruff patch. While, we express differences of opinion here, I like to think we are all also friends in a way, making our respective contributions to this community.
I would hope anyone of us, who might be going through a personally challenging time would be comfortable sharing about that.
There is so much technical expertise here and business management expertise (successes and horror stories) and I think personal concern for each other, that I feel people would open to being supportive.
Stay well.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 31, 2024 9:08:30 GMT -6
Cool and by the way, it's hard to relay in text but I don't take any of this seriously. I have a passing technical interest in audio due to previous jobs but I don't even do that any more. I could just as happily chat about car engines or TV shows over a cold one..
On a public forum even if it's a direct quote I'm aware that others read it and hopefully this topic helps someone?! It's nice to bridge the realms of technical vs. reality because many, many years ago I found it confusing personally. I can't say having the answers makes me a better mixer though .. Ultimately, it's something to chat about and if Nyquist ain't your thing okay, if it's someone else's I'll happily talk to them about it.
All good, my dude. What I meant is that I’m interested in the more practical side or making music and recordings. Probably should have added a smiley face or something. yeah all my posts in this thread I want to convey that the manufacturers, designers, and programmers should take care of all the technical things for you so you can use the equipment to make recordings and music without having to think about your signal path when you turn it on. But they mostly don’t. Sad.
|
|
|
Post by ab101 on Jan 31, 2024 11:02:42 GMT -6
I hear what you are saying Dan about having those folks take care of the technical things, and maybe Lynx included something that can throw off people, but that should be balanced with the great customer service of Lynx to assist the customer to make the best decision. And the Lynx manual for the Hilo may assist with this as well. If the attitude of Lynx was - here is our product - good luck - that would be another story. Fortunately, Lynx is not of that mindset.
We all err, but we all have great points as well! Dan - I have learned a lot from you on RGO, and I am grateful for that.
Cheers!
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 31, 2024 11:26:52 GMT -6
yeah all my posts in this thread I want to convey that the manufacturers, designers, and programmers should take care of all the technical things for you so you can use the equipment to make recordings and music without having to think about your signal path when you turn it on. But they mostly don’t. Sad. I can definitely relate, when I did a demo many years back with a certain DAW it wasn't actually compensating properly for latency even within its own internal mixer. You could send to an aux with no plugins on it at all and it would be out of phase. I didn't realise for a while after and now I'm absolutely paranoid about it, I read the purple site later on and the amount of times I've seen this issue crop up with various DAW's is unbelievable. Not only that but DSP console mixers as well, even some AVB interfaces required you to route through the DSP mixer or it would again be out of time, the manufacturers were aware of it as well.
It's not all doom & gloom, many things over the years did what I expected of them (some didn't) but on occasion I did really wish I could afford to go back to a console with well recognised tools and just use a DAW as nothing but a basic editor / tape deck.
|
|
|
Post by bubblejackets on Feb 1, 2024 0:35:07 GMT -6
Interested in what the “more transparent” conversion sounds like. Seems like it’s kind of focused on the Line in/out and maybe not the monitoring?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 1, 2024 2:53:31 GMT -6
Interested in what the “more transparent” conversion sounds like. Seems like it’s kind of focused on the Line in/out and maybe not the monitoring? Everything has to go through a converter, even monitoring playback (D / A) or we wouldn't be able to hear it. Transparent will often sound thin or boring, when we start stripping out things like harmonic distortion which is a vital mixing tool for speaker playback and general excitement enhancement things might not go the way you think. So, a quick example is let's say I play a 30Hz sine on a 5" speaker with a reproduction limit of 65Hz, what happens?.... Nothing, you wouldn't hear it, in a regular representation you'd hear the harmonics of a fundamental note. That alone shows why you'd add HD and how important it is. There's many more examples of anti-boring, we often like tape for example however that's usually just soft clipping or "saturation". You're exceeding the limitations of this specific device which ends up causing a wave to be rounded at its threshold, this is not to be confused with horrible hard clipping which happens when you exceed the limitations of a converter or even anything generally with opamps in it (even if the effect is slightly different). Anyway, before this turns into a book on audio when we mix we do a lot of things that are entirely contrary to what a converter would ever do and the funny thing here is tape was a recording medium which mangled stuff. Some liked it, some didn't.. Going further I can even understand people liking jitter, I mean we use delay's and verbs, we double track guitars for variations in timing. However the whole point of all of these technical threads are, from a logical standpoint why would you ever want your converter to do any of this? Mainly because simply put you have zero control over it, jitter isn't a beat quantized delay plugin, if something was out of time like a guitar recording you'd just re-record it right? But jitter is random so it's like shooting in the dark. Neither could you have perfectly added even or odd harmonic distortion at chosen frequencies / specific noise that suits your particular taste and works for every track. This is why we've had tools since? When did the Aphex Aural Exciter come out? Anyway the point again is you can do all of this after the fact but in a controlled manner, so again why would you want your converter doing this again? A converters job isn't necessarily to sound great, what it is meant to do is keep a signal intact so you can make it sound great. I think one of the issues vs. classical engineering is the approach, not only was old equipment technically inferior which added its own gubbins in, the main methodology of getting a "sound" was selecting hardware to interact and compliment each other. Mic's were even developed to be tracked to tape for example, nowaday's without consoles, outboard, tape or even in some cases external preamps which interface preamps are generally great but they're usually made to compliment manufacturer specs which means dead quiet, clean, no saturation, nothing.. With ITB you have to think about stuff you wouldn't in a traditional setup, latency, pan law, harmonic distortion, dither etc. and its become more of a technical approach in some area's. Coming from a console background even though I'm a techie originally I found this stuff confusing at first and it did come to bite me on occasion.
|
|
|
Post by bubblejackets on Feb 1, 2024 10:10:55 GMT -6
I found this stuff confusing at first and it did come to bite me on occasion. I find stuff confusing at all times and it comes to bite me on all occasions haha. At least I'm consistent? I'm just wondering how the new version 2 will compare to the one now. I like the original one for its monitor and headphone outs conversion. But I'm also using the line in/out for outboard loop and tracking. Unfortunately one of my line in channels has developed (I think) a noise issue that seems internal to the unit. Lynx has been great with support but we haven't figured it out yet I couldn't totally tell from their website if Lynx has changed the converters for just the line in/out or also the monitoring or what?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 1, 2024 13:39:24 GMT -6
I found this stuff confusing at first and it did come to bite me on occasion. I find stuff confusing at all times and it comes to bite me on all occasions haha. At least I'm consistent? I'm just wondering how the new version 2 will compare to the one now. I like the original one for its monitor and headphone outs conversion. But I'm also using the line in/out for outboard loop and tracking. Unfortunately one of my line in channels has developed (I think) a noise issue that seems internal to the unit. Lynx has been great with support but we haven't figured it out yet I couldn't totally tell from their website if Lynx has changed the converters for just the line in/out or also the monitoring or what? Ohh yeah, this a Lynx Hilo thread , from the specs it looks like nothing's changed on the monitor outs.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 1, 2024 14:33:26 GMT -6
I found this stuff confusing at first and it did come to bite me on occasion. I find stuff confusing at all times and it comes to bite me on all occasions haha. At least I'm consistent? I'm just wondering how the new version 2 will compare to the one now. I like the original one for its monitor and headphone outs conversion. But I'm also using the line in/out for outboard loop and tracking. Unfortunately one of my line in channels has developed (I think) a noise issue that seems internal to the unit. Lynx has been great with support but we haven't figured it out yet I couldn't totally tell from their website if Lynx has changed the converters for just the line in/out or also the monitoring or what? I think they just put the updated clock from the Aurora (N) into the Hilo. I could be wrong. I want both of them.
|
|