Advice On Audio Interface Purchase

[laxu]

I had a bright idea to add a few extra mic pres to my existing audio interface. I dug out my old Echo AudioFire 4 and plugged that into my Focusrite Scarlett 6i6 2nd gen via SPDIF.

Of course, that didn't work out because the Echo AudioFire 4 didn't happen to be configured right. Since it's Firewire, I have no way to connect it to anything to configure it.

 

[James Freeman]

Trivia;
Firewire (IEEE 1394) was standardized around 1995 and had a speed of 400 Mbit/s, very convenient for plug-n-play digital data transfer, etc.
USB 1.0 was standardized in 1996 but was slow AF with pathetic 1.5 Mbit/s, it was not until 2001 with USB 2.0 that USB reached the speed of 480 Mbit/s but by then Firewire was the go-to for a ton of studio equipment and remained so for at least another decade.
---

Hopefully USB will be backwards compatible.... forever.

 

[Sascha Franck]

Since it's Firewire, I have no way to connect it to anything to configure it.

According to some articles, FW is still working under macOS Sonoma. Possibly not for anything performance related, but you might be able to install a control panel and adjust some settings. So all you may need is a TB/USB to FW adapter and some hope for the utility software to not use anything 32bit anymore.

 

[norminal]

Anyone know if using a coaxial S/PDIF to Toslink adaptor would add additional latency? Would I be better off just going analog out of my modeler into an interface?

 

[James Freeman]

Anyone know if using a coaxial S/PDIF to Toslink adaptor would add additional latency? Would I be better off just going analog out of my modeler into an interface?

No additional latency.

 

[Sascha Franck]

And fwiw, as Zoom f***ed things up royally (their firmware installers don't even allow me to fully revert to an older version, so the unit is in fact half-bricked...), I'm in the market for a new interface, too. Fortunately, ATM I really only need a simple 2 I/O solution so I may order a Motu M2 later on. Considered a used Babyface, but as the Hi-Z input comes with an impedance of 470kOhm, I'd rather test it before buying, which seems impossible (all offers are from the other end of the country).

 

[Sascha Franck]

Anyone know if using a coaxial S/PDIF to Toslink adaptor would add additional latency?

No.
Oh, whoops, James beat me to it.

 

[norminal]

No additional latency.

So going with say, an Audient iD14 MKII and one of these, it’d still be an advantage over going into it analog?

Optical-to-Coaxial OR Coaxial-to-Optical Digital Audio Converter, ROOFULL Bi-Directional Digital Coaxial to/from Optical Toslink SPDIF Audio Converter/Adapter/Repeater

 

[mavrick102000]

Direct monitoring latency improved from the MKI to MK 2 and a slight improvement on RTL, but the HP power out you have to be careful with.

HP power is improved and and a flat response but still requires at lease 80Ohm cans to get the best listening experience up to about 150 Ohms. Any higher Impedance headphones above 150 ohms will required USB-C to USB-C connection to use 150 - 600 Ohm Headphones. USB-C to USB-A cannot supply sufficient power above 150 Ohms.

You can watch Julian Krause's review here for the details.

 

[norminal]

Any higher Impedance headphones above 150 ohms will required USB-C to USB-C connection to use 150 - 600 Ohm Headphones. USB-C to USB-A cannot supply sufficient power above 150 Ohms.

I’m using 300 ohm headphones at the moment, but I have the capability of running USB-C to USB-C.

 

[James Freeman]

So going with say, an Audient iD14 MKII and one of these, it’d still be an advantage over going into it analog?

Optical-to-Coaxial OR Coaxial-to-Optical Digital Audio Converter, ROOFULL Bi-Directional Digital Coaxial to/from Optical Toslink SPDIF Audio Converter/Adapter/Repeater

Yes, but you will have to set the audio interface to external clock, otherwise you will hear clicks.

 

[Sascha Franck]

Fwiw, just got my Motu M2. It's almost phenomenal. The two most important things:

- RTL at 44.1kHz, 32 samples buffer size (which I will easily be able to treat as "set and forget" on the new Macbook Air) is 3.5ms. That's every bit in the league of RME.

- The Hi-Z input (in fact, it combines Hi-Z and line in without any adjustment options - but that seems to work absolutely fine) is great! Helix Native sounds and responds just like the real deal. I haven't done any more scientific comparisons (maybe later), but I've played the Floor and Stomp enough to know pretty well how they feel under my fingers.

Caveat: driver-to-host communication is not correct. At the same settings, the driver reports a latency of 2.5ms - but the analog measurement via RTL utility told the real truth. There's a difference of 45 samples between reported and measured latency. As a result, there's a recording offset of the same value - something that defenitely shouldn't happen with an interface of a company such as Motu. Fortunately, Logic allows to globally compensate for that offset, but you really shouldn't have to do that (that compensation is meant to be there when using external converters the interface doesn't know anything about).
I already filed a support ticket and amazingly enough, a Motu guy got back to me just half an hour later. Let's see what they will make of this. Should be a kind of easier fix, but as a result, the reported latency numbers would be higher than before, causing existing users (unaware of the latency being higher anyway) to get all salty. We shall see...

For now, I'm a super happy camper, though. 3.5ms is pretty much stunning and the dynamic feel is great, too.

 

[norminal]

Well, I ended up grabbing a gently used Audient id14 MKII. It seems to “tick all the boxes” for me. I’ll be running my FM3 into it via S/PDIF and will use one of the mic inputs for mic/vocals. The form factor is important as the controls are on top versus the front on most other interfaces. The footprint is perfect for the space I have available. But like anything, we shall see.

Thank you for your input everyone.

 

[laxu]

I ordered the Audient EVO 8 as a local store had a small sale on it. The features seem good for <200 € price and latency seems to be similar to what I get from my Focusrite Scarlett 6i6 2nd gen. Compact form factor is a plus too. I can return it if I don't like how it functions or performs.

I felt the auto-gain feature will likely be very useful for setting up the gain level for multiple mics. The plan is to use the extra mic pres on the EVO 8 to setup a FRED mic setup with dual SM57s on my BluGuitar Nanocab (I think it records better than the Fatcab) and the AT2020 or XREF20 as either a room mic or close mic on the Fatcab.

I also picked up the Melda MAutoAlign plugin on sale. It's a track phase correction plugin that is dead easy to use and will save me a ton of time setting up mics to be exactly in phase. You just add the plugin to the tracks you want to correct, set them all to the same group (so you can use the plugin for different sets of tracks), hit analyze and it will calculate and apply the required correction.

 

[laxu]

Audient EVO 8 arrived.

The good:

• It's compact. About 1/3 the depth of my Focusrite Scarlett 6i6 2nd gen, slightly narrower and a bit taller.
• 1MOhm instrument input in front panel.
• 4 mic preamps. This is surprisingly hard to find until you get to much larger interfaces.
• The single knob and a bunch of buttons UI works surprisingly well.
• Ability to easily mute either set of outputs by holding the output buttons is nice.
  • You just hold the button for the output you want to mute. Reasonably convenient way to swap between monitors and headphones without rolling down volume knobs.
• Separate settings for 48V phantom power instead of just phantom power on/off.
• Audio levels can be controlled from MacOS. A lot of audio interfaces on MacOS will just not let you adjust volume at all, and you need to do that from the interface or its mixer software.
• Ability to have an "Artist mix" which is basically a separate level/pan configuration for outputs 3+4. Good if you want to have e.g just a metronome to your headphones while tracking.

The bad:

• Plugging in headphones mutes the equivalent output in the back. While in most cases this is what you want, I tend to leave my headphones and monitors connected and just use whichever is convenient. I'd rather select between the two instead. I'm getting around this by just using the 2nd headphone jack for headphones, exactly like I did on my Focusrite.
• Instrument input is tied to Mic/Line input 1 in the back rather than being separate, so if you plug in at the front you can't use the input in the back.
• The headphone amp is not the most powerful out there.
  • Since I use EQ correction for headphones, that tends to reduce the level further and I seem to have to turn the volume higher than I did on my Focusrite in comparison.
  • The EVO 8 being bus powered vs mains powered Focusrite is probably a factor though.
  • It still has enough range for my 250 and 300 ohm headphones to get very loud, so not that big a problem.
• The LEDs around the knob illuminate based on the output level when you are e.g playing music. I've never liked flashing lights on anything, would be nice to be able to disable that. But not a big deal since it's not like they are super bright.
• Plastic case doesn't seem the most durable thing ever, but it's not like you move these things that much...

I wish it had:

• SPDIF and MIDI.
• An extra set of 1/4" line inputs so I don't need to swap between my Hydrasynth and mics.
• Power switch.
• Separate volume controls for headphones.

The EVO Mixer software is alright, but I immediately found a bunch of little usability omissions and sent Audient some feedback that I hope they address. Mostly simple stuff like "double click pan or faders to set them back to default position (center for pans, 0 dB for levels)", or "show numeric values for knobs so it's easier to find the right spot".

Latency is not the best, but nearly identical to what I get from my Focusrite.

Compared to DAWBench results for the Focusrite Scarlett 6i6 2nd Gen, I seem to be getting a bit higher latency. Audient specs say RTL for 44100 and 96000 KHz @ 32 sample buffer size should be 5.1 and 4.0 ms, and those numbers seemed to line up with Julian Krause's review of the EVO 4. I'm about 1-1.5ms higher here.

• Might be a difference between MacOS vs Windows+ASIO.
• It could be because I have a pile of programs running and can't be bothered to quit everything just to measure.
• Rogue Amoeba's Soundsource app/driver may be adding latency even when disabled. I use it as global headphone correction on my system.

Results from RTL Utility using a M2 Max Macbook Pro 16".
Used a TRS cable from Out 3 -> In 3 to test, for no other reason than it was convenient to plug in.
Text colors just to make it easier to read the different sample rate / buffer size combinations.

Sample rate (KHz)	Buffer size (samples)	Focusrite Scarlett 6i6 2nd gen	Audient EVO 8
44100	16	5.828 ms	5.760 ms
	32	6.553	6.485
	64	8.005	7.914
48100	16	5.500	5.667
	32	6.167	6.333
	64	7.500	7.646
96000	32	5.031	5.229
	64	5.698	5.896

What it does is totally fine, I'd say even good, for a sub-200 € interface. But it sucks that the larger EVO 16 does not seem to perform any better for latency. Too bad there's nothing between the EVO 8 and EVO 16 either.

 

[Sascha Franck]

You just hold the button for the output you want to mute. Reasonably convenient way to swap between monitors and headphones without rolling down volume knobs.

Now that's cool.

Might be a difference between MacOS vs Windows+ASIO.

Yeah - as I learned from you fine folks, Apple is now onto a different driver model, even RME is only able to deliver their lowest latencies once you re-activate Kernel driver access in safe boot mode. They have a new DriverKit version for all devices, too, but they're saying themselves it's not as rock solid. So any DriverKit drivers may as well perform worse than ASIO equivalents under Windows (so far they were always roughly the same in case there were dedicated drivers for each platform and not Thesycon vs. Core Audio).

It could be because I have a pile of programs running and can't be bothered to quit everything just to measure.

That shouldn't affect latency. It certainly doesn't over here (and never did).

Rogue Amoeba's Soundsource app/driver may be adding latency even when disabled.

Ouch - if that was the case, they should be slapped.

What it does is totally fine, I'd say even good, for a sub-200 € interface.

Defenitely, especially in case you're not relying on audio software monitoring (for virtual instruments it's roughly half of it anyway, plus, if you are a keyboard hack like me, you'd quantize everything anyway...).

 

[laxu]

Yeah - as I learned from you fine folks, Apple is now onto a different driver model, even RME is only able to deliver their lowest latencies once you re-activate Kernel driver access in safe boot mode. They have a new DriverKit version for all devices, too, but they're saying themselves it's not as rock solid. So any DriverKit drivers may as well perform worse than ASIO equivalents under Windows (so far they were always roughly the same in case there were dedicated drivers for each platform and not Thesycon vs. Core Audio).

DAWBench results show the Focusrite Scarlett 4i4 4th gen having higher latency than the 6i6 2nd gen. I'm only about 0.5ms higher than the 4th gen numbers so it's possible that different driver model is the cause.

That shouldn't affect latency. It certainly doesn't over here (and never did).

Reading through RTL Utility's FAQ, they do say that apps may have an effect on it but most likely only if audio sources are playing at the same time, you have your DAW open at the same time etc.

Ouch - if that was the case, they should be slapped.

Soundsource has its own driver you need to install so it actually works, and it even supports loading AU/VST plugins so it's not insane to think that it might have an effect on latency. I just use its graphic EQ with AutoEq presets to correct my headphones because Sonarworks proved to be so damn unreliable and crashy.

Defenitely, especially in case you're not relying on audio software monitoring (for virtual instruments it's roughly half of it anyway, plus, if you are a keyboard hack like me, you'd quantize everything anyway...).

I wish I was a keyboard hack. I'm more like a keyboard caveman. It sucks to be able to play guitar and bass reasonably well, but be a total beginner on keyboards or drums, and having to somewhat start all over to learn those.

 

[Sleezy E]

Separate settings for 48V phantom power instead of just phantom power on/off.

This is a big deal if you plan on using a ribbon mic

 

[Sascha Franck]

Soundsource has its own driver you need to install so it actually works, and it even supports loading AU/VST plugins so it's not insane to think that it might have an effect on latency.

Oh, I'm sure it will - but you said it might even add some if deactivated.

It sucks to be able to play guitar and bass reasonably well, but be a total beginner on keyboards or drums, and having to somewhat start all over to learn those.

Don't even tell me...

 

[laxu]

Kudos to Audient support. They answered very quickly and pointed out that you do have shortcuts for resetting pan and fader levels. It just requires an Option/Alt click on them. Also answered some of the other things I suggested on whether they were possible or not.

As expected, "headphones mute back outputs" is hardwired so nothing can be done about that.