Testing latency of USB Audio Interfaces

[djbclark]

I'd like to test latency of a range of USB Audio Interface DACs.

So questions:

(1) Is there a canonical test or series of tests that do not rely on subjectivity to use to do this? Eg when people say "this cut X time off my latency" how exactly are they concluding this unless it's so bad it's super-obvious? Must run on Mac OS X 10.15, but running on Android as well would be great.

(2) Any suggestions on DACs to add? They must either be under $20 or available from a place with easy returns like Amazon.

I'm planning on testing at least these 5:

Steinberg UR12 USB Audio Interface, $110.38
https://www.amazon.com/Steinberg-UR12-U ... B00QY4RLRQ
Someone mentioned the UR22 getting rid of latency problems in their setup, this seems like it should be the same thing just with fewer ports.

Just jumping in here from another thread to crow about the Steinberg UR22-C I just picked up. [...] But the biggest shocker using this new unit is that it pretty much disappears all the latency when recording!

SAMSUNG EE-UC10JUWEGUS USB-C to 3.5mm Headphone Jack Adapter, $13.88
https://www.amazon.com/dp/B07VF6NSZZ
Bought this randomly before, seems to take all HD formats no prob.


MUSEHIFI M1 Smart, $12.99
https://www.aliexpress.us/item/3256803429980061.htm
Was in a review where it did okay, and this is a amazing price for it. The power in port also looks like an aweome feature for smartphone use.


MUSEHIFI M1, $10.94
https://www.aliexpress.us/item/3256803370659232.html
Was in a review where it did okay, and this is a amazing price for it. Tiniest offering I've found.


Apple USB-C to 3.5 mm Headphone Jack Adapter, $8.48
https://www.amazon.com/gp/product/B07K25P3N1/
It doesn't go above 48hz but latency could still be good, plus many probably already have these so could be point of comparison.

Some of this stuff is coming from AliExpress so prob won't be able to complete the chart until mid-Dec.

This is the closest thing I found to the kind of review I'm looking for: https://www.youtube.com/watch?v=1PO4sNDhvvo

[jacko]

There is a thread (possibly a couple) on the vdrums forum that goes into this in some detail. have a look also on the gearspace forum

[djbclark]

I couldn't find any pre-existing threads, but started a new ones:

https://www.vdrums.com/forum/advanced/d ... omparisons

And at gearspace:
https://gearspace.com/board/audio-stude ... st16254204

[LooseSends]

Can't wait to see what you come up with!

In my case I was using a very cheap DAC (Behringer UCA222) before pulling the trigger on the Steinberg, and the difference in latency was audibly noticeable. However, it might not have been exclusively an A/B switch out as there's some tech in the Steinberg hardware/software that can cheat the latency somehow. At least that's what they claim. Another factor is that I moved to a much faster computer for hosting, although that was a week or so after getting this.

That's why I'm curious about some actual numbers, as my experience might be subjective. At the end of the day, all I can say is that anything I put through the Steinberg has effectively no discernable latency compared to my previous setup, and I'm very attuned to small latency differences. The midi transmission distance is even longer because that faster computer is in another part of the house and not right beside the DAC, not that it likely matters that much.

[mprinz]

Hi,

but, why these Samsung etc. devices.
Do they support ASIO?

I think it's basically the driver incl. buffersizes ... that the hardware.

I am quite lucky to be able to use a M-Audio Fast Track Ultra from 2008 that let
me use the eDRUMin10 with SuperiorDrummer3 with a latency of 7.5 ms (64smpl@44.1kHz).
And I just check some Motu and Audient devices and ... Nope, not faster.

[Rob]

I upgraded from an M-Audio 1814 to an RME Babyface. The 1814 did a fairly good job, and while the RME idoes allow me to go a bit lower in terms of latency, the biggest difference is improved sound quality. Very noticeable in terms of stereo imaging and low frequency reproduction.

[mprinz]

Ohhhhkaaaay.

These are ... créme de la créme .

But 800$ ... I actually was checking ebay for a used Babyface... But still a lot of money.

PS: Rob, do you have an idea for my pos sens not going up to 127 from the "Erratic positional sensing" thread.

[mprinz]

I upgraded from an M-Audio 1814 to an RME Babyface. The 1814 did a fairly good job, and while the RME idoes allow me to go a bit lower in terms of latency, the biggest difference is improved sound quality. Very noticeable in terms of stereo imaging and low frequency reproduction.

Hi Rob,

finally, me too.

Damnm this thing is so good.
Exactly, sound quality. The low frequencies .. I noticed that right away.
And that TotalMix Software is awesome.

I also grabbed a used Focusrite Saffire 40 Pro and connected it via ADAT to get some more I/O.

And - it might be placebo, but I feel like I can hear the better performance in terms of output latency.
Great!