Which DAW has lowest latency when tracking vocals? (Without plugins inserted)

[simplersounds]

Assuming all other factors are removed and one is using a lower latency soundcard like RME, which DAW has lowest latency when tracking vocals without plugins

[echosystm]

Logic.

CoreAudio is superior to ASIO and WASAPI, most of the time.

Sound card hardware has a trivial impact on latency. Latency is mostly a software issue, with drivers being the main determinant.

CoreAudio is more efficient in its implementation, but you are also (generally) not subjected to whatever half ass drivers a given vendor has produced.

That said, if you are only recording, then you should probably be using direct monitoring. ASIO supports this and (AFAIK) CoreAudio does not. I would expect any non-broken DAW to implement this flawlessly and they should all be equal. I can't speak to what happens in the real world though. You'll probably need to test it yourself.

[jancivil]

Well, my experience with RME on a Mac was, it was easily better drivers than the onboard thing, but being written for ASIO it is not so optimal for Core Audio and was not the experience on Mac it is under Windows OS. Also, until relatively recently Cubendo suffered under Mac OS similarly. So your 'sound card' drivers are 1) not created equal and 2) perhaps not truly platform-agnostic. Confer Windows/ASIO vs Mac/Core Audio.

Most of the problematic latency is hosting plugins, and particularly running them in the same process as the DAW host. So most DAWs with no plugins to worry about should perform as well as your drivers, given the OS and particular host are conducive and agreeable.
Absent the information, what OS and so forth, the question is not fully answerable.

After years and years of struggling (I'm not into Logic), finally I have decently low latency (low enough to drum predictably on time) and it's down to my audio interface, which uses a technology they call Direct Memory Address (much more technical writing than I'll do, mind) which takes a load off CPU. Presonus Quantum is the name of the product line.

[echosystm]

DMA is more about how the sound card is connected to your computer than the sound card itself, but that is a fair point. I forgot about stuff like that, because everything is USB nowadays.

DMA isn't possible over USB, at least not in the same way as PCIE/Thunderbolt/FireWire. USB is packet based and will always eat some CPU. Back in the old days, everyone used PCI and FireWire, because of this. CPUs and USB eventually became fast enough that folks stopped caring about it as much, but there is no question that the other connection types are better.

[jancivil]

It uses Thunderbolt 3, which is a type of USB3. The specific implementation doesn't happen outside of Thunderbolt 3.

The Quantum-series proprietary Thunderbolt driver was written from the ground up [...]

The Quantum driver utilizes the bus master Direct Memory Access (DMA) to transfer audio data directly to and from the main memory without using CPU processing and the overhead that would come with it. DMA supports 64-bit addressing on systems with 4 GB or more of RAM. A companion DMA engine was also implemented in the Quantum hardware to read digital audio data directly between the ADC/DAC and the computer memory, one sample at a time.

This DMA is certainly a property of this device, per se.
https://www.presonus.com/products/quantum-2626#Speed

[Uriel Anthony]

The Presonus Quantum is probably the fastest audio interface to date hands down.

Thunderbolt 3 is not a type of USB3 though, it's just that they both uses the same USB-C plug/socket interface. So if you plugged in a USB device into such said port, the signal will be routed to a USB host controller chip on your motherboard and if you plugged in a Thunderbolt device then the signal is routed to a Thunderbolt controller chip instead. Because of this it has resulted in a lot of confusion regarding hardware compatibility.

[echosystm]

It uses Thunderbolt 3, which is a type of USB3. The specific implementation doesn't happen outside of Thunderbolt 3.

The Quantum-series proprietary Thunderbolt driver was written from the ground up [...]

The Quantum driver utilizes the bus master Direct Memory Access (DMA) to transfer audio data directly to and from the main memory without using CPU processing and the overhead that would come with it. DMA supports 64-bit addressing on systems with 4 GB or more of RAM. A companion DMA engine was also implemented in the Quantum hardware to read digital audio data directly between the ADC/DAC and the computer memory, one sample at a time.

This DMA is certainly a property of this device, per se.
https://www.presonus.com/products/quantum-2626#Speed

Thunderbolt is not a type of USB. Thunderbolt is PCIE, over a cable. That is what allows DMA to be possible. Writing a driver for a PCIE/Thunderbolt device like this and not using DMA would be silly. The marketing would make you think that something revolutionary is happening, when it is not. DMA has existed since the 80s. This is how all decent PCI/PCIE audio interfaces work. Class compliant FireWire sound cards work in a similar way on MacOS, for "free" in the sense that no vendor driver is required. ASIO drivers are always required on Windows, because WASAPI is still inferior.

[saltbeatz]

Biggest impact on latency, in my experience, has been hardware. With an RME babyface pro fs I can do realtime recording with plugins without issues on multiple daws. Never noticed a real difference DAW for DAW; however, studio one's "dropout protection " is cool as it turns off plugins if you go to record with monitoring on.

[jancivil]

"This is how all decent PCI/PCIE audio interfaces work. Class compliant FireWire sound cards work in a similar way on MacOS, for "free" in the sense that no vendor driver is required."
Not only is this bulllshit, there is no actual sense to the assertion.

T3 is known as a superset of USB-C. It is not exactly the same thing as that, and it is not a subset of it. It uses the Universal Serial Bus, I'ma call it USB.

As a point of empirical fact, I have never had latency this low for the things I do, not even remotely in the same ballpark. I don't know who benefits from such broad theoretical assertions. I'm actually seeing it asserted that all PCI type audio interfaces use the technology Presonus claims written 'from the ground up'? A rather extraordinary assertion.
Very low latency with Cubase (using virtual instruments heavily) under OSX is passing difficult, I've been at this since 2006. I'm not certain why, but this is the only thing that has sorted me in 15 yrs.

This is_not FW, the "free" status is entirely irrelevant, since for certain it uses proprietary drivers/is demonstrably not plug-and-play.

[ghettosynth]

I prefer to just use direct monitoring and the EQ/Dynamics/Verb built into my MOTU interface for vocals. This way tracking latency does not impact monitoring latency. That won't work for all use cases, but, it's something worth considering when thinking about buying an interface. If your playback involves heavy plugins then you may not be able to use the latency that is available to you. Direct monitoring mitigates this, albeit, at a cost.

A similar discussion on the other site has a nice discussion of some of the issues in detail with respect to contemporary interfaces as well as related assertions re:marketing speak.

https://gearspace.com/board/music-compu ... e-124.html

emphasis mine

From my point of view, this absence of any hardware input monitoring and audio processing at Quantum is its biggest difference compared to others. And it should be involved in decision about its purchase.

For some users and use cases, it doesn't matter and straight I/O to DAW can be even preferred solution (you surely know some users having instant headaches from any interface control software ). Similarly when some outboard monitoring (like external analog mix) is used for tracking or user primarily record using amp sims or VIs (and thus DAW software monitoring is always necessary).

Other users (mine included) like to have the option to switch on hardware input monitoring with constant latency regardless of buffer size in DAW, because there can be situations, where your computer simply won't cope with load associated with short buffers there.

Regarding the timeline of "DMA technology."

From the wikipedia page for the 8237 DMA controller:

Intel Corporation, "Microcomputer Components: New Intel 8237 DMA Controller provides a 5 MHz DMA answer for 8088 and 8085A-2 based systems", Intel Preview, May/June 1979, Pg 9.

[trusampler]

I'm using Presonus Studio One on Win10 with Antelope Audio Discrete 8 Synergy Core tracking vocals at 8 samples via Thunderbolt/ as well as their USB driver, makes no difference with this interface.

[ghettosynth]

I'm using Presonus Studio One on Win10 with Antelope Audio Discrete 8 Synergy Core tracking vocals at 8 samples via Thunderbolt/ as well as their USB driver, makes no difference with this interface.

I spent way too much time debating whether or not I wanted to jump on that train. I also considered presonus' rack mount mixers.

I don't need new hardware.
I don't need new hardware.
I don't need new hardware.

Ok, it's passing now...

[Amberience]

As far as I'm aware, CoreAudio has more inherent latency than ASIO, because there are more calls to the kernel per sample-block than ASIO.

But it's not a huge difference. Mostly you'll need good hardware to achieve low latencies. Just make sure you avoid the Thesycon driver on Windows!

I'm currently using a Universal Audio Apollo x8p, which doesn't give me the lowest latencies around. But I can use some decent plugins. However I've been working on finding replacement native plugins over the last 5 months or so, because I upgraded my machine to an AMD 5950X, so I have a lot of CPU juice on hand.

So I'm thinking that I will switch my audio interface. Antelope Audio, RME, or Presonus, seem to be the ones to go to for low latency performance.

[jancivil]

I just don't know enough to grok why Cubase on Mac needs the extra latency, I had a guess years ago it was something to do with its use of a thing called ASIO2CoreAudio dot bundle (which I noticed in a crash log). Or was it CoreAudio2ASIO? IE: a translation. But who knows.

A comment - probably here - that took on a life of its own, got parroted incredibly widely on every known thing, people still argue it and a lot of people took it as fact which the people that don't believe it really hate on.
IE: a classic.

[ghettosynth]

Also, many things microcomputer were not new ideas really.

DEC co-founder Kenneth Olsen led the circuit design for the transistorized TX-0 computer in 1956. The new company began by designing and selling transistorized circuit modules for building digital systems; it subsequently applied these techniques and products to its own computers, beginning with the PDP-1 in 1960.

With its 18-bit, 4,096-word memory, DEC’s PDP-1 was priced from $85,000-100,000 for a standard installation. Its model name, for “Programmed Data Processor,” helped especially government buyers circumvent long procurement processes intended for mainframe computers. Designed to run on standard power in a typical office, the PDP-1 contrasted with mainframes in its installation requirements, small size (17 square feet), and user-friendly interface. It was designed to connect to and support a variety of input/output (I/O) devices for high-data-rate, real-time uses. For example, it had a multi-level interrupt capability to respond to real-time devices; nearly half of the 53 PDP-1s sold went to ITT for “message switching” interfaces with teletype lines. Drawing on the TX-0, PDP-1 introduced direct memory access (DMA) ports so that data from external devices could be fed directly into memory with minimal demands on the central processing unit (CPU). [/quote