Lets have real talk about audio interfaces

Anything about hardware musical instruments.
Post Reply New Topic
RELATED
PRODUCTS

Post

sqigls wrote: Sun Jan 06, 2019 4:20 pmanyway Woz thought long and hard about it all a billion years ago, one reason once you go mac, you never go back. In my honest and hard-earned user experience.
That's the funniest thing I have heard today. I spend 8 hours a day on a MacPro at work and I can honestly say I have more problems in a day with those various pieces of garbage than I have with my Windows 10 laptop in a year. Because I work as a graphic artist, people normally assume that I would be a Mac guy but once I explain to them that I am a PC guy precisely because I have to use a Mac every day at work, they start to understand.

I was first exposed to (supposedly) high-end Macs in about 2003 but there has never been a time since then that I have seen any evidence that they were one iota better than an equivalent PC in any way -not faster or more powerful, not easier to set up or use and nowhere near as reliable. You literally could not give me a Mac.
NOVAkILL : Asus RoG Flow Z13, Core i9, 16GB RAM, Win11 | EVO 16 | Studio One | bx_oberhausen, GR-8, JP6K, Union, Hexeract, Olga, TRK-01, SEM, BA-1, Thorn, Prestige, Spire, Legend-HZ, ANA-2, VG Iron 2 | Uno Pro, Rocket.

Post

Dasheesh wrote: Wed Dec 26, 2018 3:43 am Honestly, if some of you "talented" programmers want to make money, design a great audio driver for windows.
What's wrong with ASIO4all? I use it all the time and it never misses a beat (literally). I think I run it at 512 samples.
rod_zero wrote: Wed Dec 26, 2018 7:40 am Just go with RME and don't buy another interface in two decades, you will question yourself why you didn't get one sooner.
I could say the same about the $250 Edirol interface I bought in 2003. My bandmate still uses it and it works at least as well as any of the half-a-dozen I've bought since. An RME Babyface costs $120 more than my laptop did, I know that in two decades I'd still be cursing myself for wasting so much money on something that didn't make any kind of improvement to anything.
JCJR wrote: Wed Dec 26, 2018 2:46 pmBut latency has to be unreasonably short before it feels right to me playing thru a softsynth in realtime. IMO 10 Ms feels too sluggish for that. For me needs to be 5 Ms or less before it feels right and even modern fast computers seem to have trouble with that because windows and macos are not "hard realtime operating systems".
Interesting that you'd say that because, many years ago, one of our most esteemed members with a basement full of hardware synths did some testing and discovered that most hard synths have latency in the order of 25ms. To be fair, your audio card's latency is not the only latency in the system but when I run my Zoom at 5ms, the total system latency that my software reports is around 18ms, which means it is faster than most hardware synths. On stage we tend to run at around 8ms, 30ms or so total, and it's fine for both me and my band-mate, who triggers drums from my laptop from his Octapad.
deastman wrote: Wed Dec 26, 2018 5:18 pm If you don’t want to pay RME prices, I’ve always found MOTU to be really solid.
Does MOTU stuff even work on PC?
yellowmix wrote: Wed Dec 26, 2018 7:59 pmThere's an audio interface latency chart here: https://www.gearslutz.com/board/showpos ... count=2186
There is zero explanation of how those results were obtained, making it completely meaningless.
tehlord wrote: Thu Dec 27, 2018 12:18 amMy RME had exactly half the latency of anything else I've owned (15 or so audio interfaces in the last decade) for the exact same buffer setting. HALF.
Don't be ridiculous, The buffer is what causes the latency. i.e. 1024 samples at 44.1kHz = 23ms, which is the latency you will get.
Drivers never produced any pops and/or clicks, ever.
Never crashed or failed, ever.
That also describes every interface I've ever used (and I have used quite a few).
dellboy wrote: Thu Dec 27, 2018 4:03 pm Could someone knowledgeable explain this whole latency thing.
I have always been happy with 10 ms, and yet people here are saying they need 2 - 4ms.
On a live stage musicians 3 meters apart experience 9 ms of delay.
Whats the truth about latency ?
The truth is that until they could read the numbers, nobody knew it existed. It's is completely in people's heads and safely ignored. As I pointed out above, most hardware synths have latency above 20ms and nobody has ever noticed or complained.
JCJR wrote: Fri Dec 28, 2018 8:07 pmYa, sound travels about 1 foot per ms or about a meter per 3 ms. However virtually all "real instruments" guitars, winds, strings, drums, except for oddballs like pipe organ, are within a meter of the performer, from 0 to 3 ms delay.
I reckon a lot of electric guitar players would be 3m or more from their stacks on a decent size stage.
NOVAkILL : Asus RoG Flow Z13, Core i9, 16GB RAM, Win11 | EVO 16 | Studio One | bx_oberhausen, GR-8, JP6K, Union, Hexeract, Olga, TRK-01, SEM, BA-1, Thorn, Prestige, Spire, Legend-HZ, ANA-2, VG Iron 2 | Uno Pro, Rocket.

Post

Dasheesh wrote: Thu Dec 27, 2018 2:49 am zoom is the only one I've seen taking advantage of USB 3 and there have been zero complaints after the initial release efforts. They seem to have in house drivers as well. Supposed to be sick low latency. I'm unsure what AD/DA converters they use but It's been what I keep coming back to. I just don't want to pop for it, have it discontinued, realize I got sold again, and then be in the same situation I'm in again... Gawd I miss echo.
Just curious, but which Zoom product were you looking at? My U24 was initially something of a revelation - such a clean sound and stupidly low latency when I bought it, 2ms was easy. But since I got a new laptop, it's lost some of its sheen. It won't run smoothly below about 8ms, which is still fine but disappointing after what I'd experienced before. I also really like the way it has been designed with portability in mind. nothing sticks out of it and most controls are protected from accidentally being changed in your bag. You can even run it on batteries. It is a very well thought-out product that I really rate.
NOVAkILL : Asus RoG Flow Z13, Core i9, 16GB RAM, Win11 | EVO 16 | Studio One | bx_oberhausen, GR-8, JP6K, Union, Hexeract, Olga, TRK-01, SEM, BA-1, Thorn, Prestige, Spire, Legend-HZ, ANA-2, VG Iron 2 | Uno Pro, Rocket.

Post

On a Mac, MOTU rock and last forever.

On pc, I'm sure you've got some great solutions too but don't know or care.

Post

BONES wrote: Thu Jan 10, 2019 5:49 am
yellowmix wrote: Wed Dec 26, 2018 7:59 pmThere's an audio interface latency chart here: https://www.gearslutz.com/board/showpos ... count=2186
There is zero explanation of how those results were obtained, making it completely meaningless.
It's in the chart itself. RTL. Round-Trip Latency. Send a signal out, record how long it takes to get back in. Anyone can perform this, as long as they have an interface with an in/out and a long enough cable with the appropriate connectors.

Post

Send a signal out of what? No methodology, just a big list of meaningless numbers. Possibly good for comparison but ultimately useless for making any kind of informed decision.
NOVAkILL : Asus RoG Flow Z13, Core i9, 16GB RAM, Win11 | EVO 16 | Studio One | bx_oberhausen, GR-8, JP6K, Union, Hexeract, Olga, TRK-01, SEM, BA-1, Thorn, Prestige, Spire, Legend-HZ, ANA-2, VG Iron 2 | Uno Pro, Rocket.

Post

BONES wrote: Thu Jan 10, 2019 6:24 am Send a signal out of what? No methodology, just a big list of meaningless numbers. Possibly good for comparison but ultimately useless for making any kind of informed decision.
I don't know if you're being intentionally obtuse when I have to spell out everything. You send a signal out of your interface and then right back in to the same interface. You can then measure the sample and time delta.

This is a "round trip latency" test. It's the most objective test possible because it includes latency not reported by the driver/software, such as that occurring in the A/D and D/A stage.

Post

Here's a SOS article that explains a lot of it: https://www.soundonsound.com/techniques ... ncy-part-1

It's from almost 20 years ago and the technology has improved but the fundamentals are the same.

Post

This article is totally unrelated to the tests on Gear Slutz. it is just talking about latency. I want to know how the test was conducted, not what latency is. How is this not completely obvious?
yellowmix wrote: Thu Jan 10, 2019 6:33 amI don't know if you're being intentionally obtuse when I have to spell out everything. You send a signal out of your interface and then right back in to the same interface. You can then measure the sample and time delta.
This is a "round trip latency" test. It's the most objective test possible because it includes latency not reported by the driver/software, such as that occurring in the A/D and D/A stage.
Where does your interface get this signal from, it is not capable of making a signal of its own you know, and how do you measure it? Does it need to be connected to a computer for this test or is there some specialised piece of equipment for doing it? How does it relate to the drivers and their settings for my laptop?
NOVAkILL : Asus RoG Flow Z13, Core i9, 16GB RAM, Win11 | EVO 16 | Studio One | bx_oberhausen, GR-8, JP6K, Union, Hexeract, Olga, TRK-01, SEM, BA-1, Thorn, Prestige, Spire, Legend-HZ, ANA-2, VG Iron 2 | Uno Pro, Rocket.

Post

BONES wrote: Thu Jan 10, 2019 6:47 am This article is totally unrelated to the tests on Gear Slutz. it is just talking about latency. I want to know how the test was conducted, not what latency is. How is this not completely obvious?

Where does your interface get this signal from, it is not capable of making a signal of its own you know, and how do you measure it? Does it need to be connected to a computer for this test or is there some specialised piece of equipment for doing it? How does it relate to the drivers and their settings for my laptop?
An audio interface does not produce sound on its own. Yes, you use a computer attached to the audio interface. You can use any combination of software you want to run this test, it literally does not matter. The goal is to send a signal, any signal, out from the interface. There is a cable attached to this output. The other end of this cable is attached to an input on the exact same interface.

You are recording that input. Whatever software you're using, you can calculate the difference between when you sent that signal to when you get it back. That is the round trip latency.

Are you asking for specific directions, as in exactly what software to use, and what audio signal to send?

Post

Exactly what I was getting at. Without knowing what sort of set-up was used, or even if the same set up was used for every test, it's meaningless. For all we know his favourite interface was stuttering all over the place to get the result and all the others had a big enough buffer setting to allow for smooth playback. I'm not for a moment suggesting that is the case but a paragraph or two about how the test was conducted wouldn't hurt and might give the results some context.

For the record, I don't care. I gauge latency with my fingers and ears and that's all the accuracy I need. Whether it's 2ms or 20ms is irrelevant to me, as long as it works.
NOVAkILL : Asus RoG Flow Z13, Core i9, 16GB RAM, Win11 | EVO 16 | Studio One | bx_oberhausen, GR-8, JP6K, Union, Hexeract, Olga, TRK-01, SEM, BA-1, Thorn, Prestige, Spire, Legend-HZ, ANA-2, VG Iron 2 | Uno Pro, Rocket.

Post

There's a link to the DAWBench site at the bottom of that Gearslutz post that has all the details.

And weird flex, but okay.

Post

Hi Bones. When I read those GS links, follow them deep enough and you find complete description of the tests and methods.

That sound on sound article looks quite good and using a daw along with the mutant wired midi cable is among the best simpleat ways to do it.

In the (old) article he measures M1 latency and jv2080 both less than 5ms which is in the same ballpark of earlier synth testing I did a few years earlier using custom hardware and software on an old toaster Mac. That was before home computers could do audio. :)

Maybe modern HW synths have got horribly 25ms slow. Dunno. Too lazy to test. It don't matter how slow something is if I don't own that model it ain't my problem to worry about. :)

Post

These weren't new synths, these were mostly vintage and often analogue and Scott Solida is someone whose word I would take over pretty much anyone else's at KVR. The thread is probably 15-16 years old now but you might be able to find it if you are interested.
NOVAkILL : Asus RoG Flow Z13, Core i9, 16GB RAM, Win11 | EVO 16 | Studio One | bx_oberhausen, GR-8, JP6K, Union, Hexeract, Olga, TRK-01, SEM, BA-1, Thorn, Prestige, Spire, Legend-HZ, ANA-2, VG Iron 2 | Uno Pro, Rocket.

Post

BONES wrote: Thu Jan 10, 2019 5:49 am
rod_zero wrote: Wed Dec 26, 2018 7:40 am Just go with RME and don't buy another interface in two decades, you will question yourself why you didn't get one sooner.
I could say the same about the $250 Edirol interface I bought in 2003.
Isn't Edirol owned by Roland? You're able to get up-to-date drivers for your Edirol interface from 2003, on the latest operating system (Windows 10 or OSX)? That's pretty incredible. As far as I knew, RME is the only company in existence who has this claim to fame. :party:

I'd still pay the money for the RME. Babyface Pro FTW! :lol:
Win 10 | Ableton Live 11 Suite | Reason 12 | i7 3770 @ 3.5 Ghz | 16 GB RAM | RME Babyface Pro| Akai MPC Live II & Akai Force | Roland System 8 | Roland TR-8 with 7x7 Expansion | Roland TB-3 | Roland MX-1 | Dreadbox Typhon | Korg Minilogue XD

Post Reply

Return to “Hardware (Instruments and Effects)”