The rme panel / software itself doesn't go lower than 64 @ 96khz, so I would assume that it's a "limitation" of the ufx+. Might be different on osx?

I'm genuinely curious here, if Apollo's have 0 latency monitoring, why would that be an issue for tracking? As stated by someone, the Quantum's are blazing fast but still, you need a hardware chain for monitoring to achieve 0 latency ( or perceived 0 latency ).

I've used the X8 before, currently use an X16 as I mainly track with hw pre's but like the HW and use the DSP while mixing. Actually I use the DSP reverbs for monitoring while tracking too sometimes...

Like everyone else I'd like extremely low RTL too, but in my real world usage, either for tracking or mixing, I'm not really noticing any issues where the Apollo's are at...

Cheers!

What Apollo calls 0 latency monitoring they are describing the signal flow of passing the live recorded input straight back to the performer’s headphones rather than going thought the computer and then to phones as it’s the trip through the computer that causes the latency - and in Apollo’s case it has plenty of RTL IF you monitor that way. Any already recorded music the performer is playing along to comes from the computer and is mixed with the live input - but the net effect is the performer hears no latency. This approach is generically called input monitoring or in TV it’s sometimes called E to E (end to end)

This signal flow is what you have been able to do with almost every audio interface of the last 20+ years if you wanted to and even kin analog studios before that. Downside is you have to deal with 2 layers of software - which is less practical and in systems that use control surfaces this becomes more of an issue as most of those systems won’t control 2 software layers at the same time.

The whole point (or at least much of the point) of having a really great performing low latency interface is so you don’t have to bother with all the BS I speak of above. You create your cue mixes in the same software you are recording into and because the RTL is so low it doesn’t matter that the performer’s live input goes through the computer - and thus we come to pretty much the whole point of this database - an accurate record of which audio interfaces you can do that with.

This kind of cue mix signal flow is how it was always done in fully analog studios for decades. The performer’s headphone mix was derived from a pre fader aux on the desk coming from the tape return channels. This approach offered one further benefit. It was safer to actually hear the recorded audio AFTER the actual tape machine rather than before (input monitoring) because if the recorder glitched at all you would actually hear it.

Kind regards

Starting to wonder whether my pursuit of low latency was really worth it. I just streamed an ableton jam that's like 4-5 song bones I can move between and I've got a guitar and vox mic setup for a few of them. I didn't even notice until I was done that my buffer was set to 512.... My brain totally adapted to it and the stream went great. Oh well

Hence their development of LUNA, which becomes a single layer of control. Apparently UA had approached other DAWS about tighter integration but they didn’t get any takers.

That’s interesting - but ultimately UA is in the business of creating a system that has people want to adopt their plugin platform and hardware. That platform necessarily needs to use the input monitor system as it is the only way they can record with their own plugins with minimal RTL - something they see as a big feature to attract new customers I am guessing.

I’ve been a big user of UAD plugins for mixing/ mastering as far back as UAD-1 and continue to mix with these plugs about 90% - but recording through plugins - UAD or otherwise is just not something I am interested in doing thus my preference for simply getting the best RTL performance I can to meet my goals.

Currently I am still sporting an original RME UFX via USB with 2 x Cranborne Audio 500Adat racks for much better sounding converters and due to newer converters shorter converter latency than the built in RME ones. 2.9ms compared to 3.3ms from the rme ones. I can record a whole live band with 4-5 stereo headphone mixes all derived from PFL sends coming from inside Logic Pro on 32 sample buffer on a 2012 5,1 mac pro with no issues.

Does anyone have any measurement of the UA Apollo Solo (USB version) on Windows? It's a USB option.

I think I remember the USB Twin did not perform as well as the Thunderbolt Twin.

Just wondering if it's "safe" to buy the UA Apollo Solo USB rather than the Thunderbolt version (same as the old name Arrow interface in a darker grey.)

I did exactly that this past week (HDSPe MADI FX + M-1610 Pro). I'll see if I can get some Oblique RTL numbers up soon.

Oh that be cool, thanks bud, that set up is my dream machine for
now but getting some RTL data could make for good confirmation.
RT

This was with the RME HDSPe MADI FX and RME M-1610 Pro, so ~1.98ms RTL at 96kHz + a buffer size of 64 samples, in 96k frame mode, with the routing for the test channels done via patchbay.

Not sure how much other connectivity details matter, but for the sake of completeness, the MADI connection was via RME SFP module (results using a Cisco GLC-FE-100FX SFP module were also identical), with word clock via BNC with the M-1610 as master. The M-1610 was also set for +24 dBu mode I/O.

According to the manual, the latency of the M1610 AD at 96kHz is 5 samples, the DA is 5.5 samples, and MADI I/O adds 4 samples.

Hopefully this information helps someone out there. If there are any requests for additional measurements, I can make time for them this weekend.

That looks really good m03 but not all that surprising, thanks for your efforts.
Think running it through the Enet>MADI>PC would make any difference?

Do you mean AVB? I wanted to experiment with running an AVB compliant Ethernet card and using something like the Hono AVB virtual sound card, but it looks like Audio Science just discontinued the Hono VSC software as of a few days ago.

Not sure what the other options available on Windows are going to end up being...maybe RME has plans for an AVB PCIe NIC + software stack?

Thanks for sharing those figures, very interesting.

I suspect the patchbay may be adding some additional delays.

190 Samples @ 44.1 is 4.308ms , so if we compare that to the HDSPe AES/ADI-8 QS numbers which are 2.947ms ( 130 samples ) @ 032/44.1 , there is an additional 60 samples.

The ADI-8 QS has around 12-14 samples respectively for I and O , so I would expect the HDSPe MADI FX / M-160 to be lower as the later AD/DA has respective lower latencies than the QS.

For further reference, the new HDSPe AIO Pro @ 44.1/032 delivers 2.494ms ( 110 Samples ) RTL

Is it possible for you to test at 032/44.1 for a direct comparison?

What I meant and if it's possible, could you get an RTL while using, yes, AVB back through
the PC with MADI and out. Might need an AVB device and maybe a switch so you can set
up an I to O, be it digital or analog, don't think doing it within routing will make a difference.

I suspected that there may be something about how I have things set up that is skewing the number slightly high. Is there an "official" established test procedure posted somewhere in the thread that I can refer to for optimal statistics gathering?

There is a user Guide for the RTL Utility at the Oblique website :

You can View/Download : Here

Simply directly connect a cable from an output of the M-16 DA to a corresponding input of M-16 AD, ( TRS might be easy if the D-Subs are connected to the patchbay) , assign that in the RTL Utility and measure directly sans patchbay.

Where can I find the compiled list of ratings so far? I can't seem to find the latest one with the latest interfaces on it.

Hi all,

i have read 70 pages but i'm not sure on the conclusion and i try to post a direct question.

I use Mac and windows daw's but necently i want to go away from apple and return in windows 100%. i don't like apple anymore... m1,m2,m3,m4,m5 .. i want a WS not a smartphone.. the new Big Sur is horrible and the price is out of any reason.

i have a TB3 capable motherboard ( Gigabyte z490 xtreme ). This board have 2 TB3 ports integrated on chipset/motherboard, no card required.

i have to admit that i'm really tempted to buy an Apollo x8 because UA hardware, DSP, exclusive plugins etc and use it via native full Tb3 with all my daw like cubase, presonus, fl studio, protools ) depends on the project , for example for some EDM i use FlStudio or LIVE. I'm interested in some UAD Plugins and a future Softtube Console 1 addon on my setup.

The problem is that i use 70% virtual instruments 30% Audio for vocals, guitars, obviously samples, sometimes real bass or some drums, depends. I know that the UA Apollo X is a very good interface , ad/da and pres are very good and i like to own one but, worth the price in ( my case ) where my primary use is WITH virtual instruments ??.

My questions :

Anyone have real life experience on how Apollo x8 perform in latency for virtual instruments ?.
Is Presonus QUANTUM TB2 interface faster tahn apollo x8 for use with VI ?.

my tipical project:

40 audio tracks ( at least 4 fx in insert )
12/16 buss ( for channel strips8 , reverbs, group comps etc )
at least 16 Virtual instruments ( a LOT of layers x elements ).

i'm not interested in working @ 2billions of Khz.. for me 44.1 / 48Khz is more than enough.

Can i hope in performance like , at least , sub 2ms when i reach 55% of progression in a project like this or this interface start to pops, clicks and struggle just with 5/6 instruments @ 128 / 44.100 ?.

i like UAD plugins expecially tape emulators, comp and channel strips but if this interface is not usable for realtime virtual instruments or i have to play instruments without any inserts ( native, obviously ) and at 10% of progression on the project i have to rise the buffer size to 128/256 ... for me is completely useless and i still with my actual Asio4all 70 bucks USB interface.

Apollo x8 ?
Quantum ?

I like Presonus Quantum because i think that the price is very good and the deep integration with studio one... but .. it's TB2 :(((((... require adapters, converters, specific cables... is 4 year old and probably presonus have some new interface for NAMM 2021 available at least in april and i don't have the time for this. If i have to change interface i have to change it now.

One of my store sell an Apollo x8 boxed.. for just 1750 tax and shippings included. 790 for quantum .

Thank you.

Someone just commented a few days ago that Apollo output latency is pretty decent, so it should be fine for virtual instruments other than ones with audio inputs (guitar/bass.) Probably fine for drums and synths.

Apollo is a massive rabbit hole and money absorber. Your apollo system and the plugins you want will end up costing you twice or more money than the Presonus Quantum system, with comparable plugins that are native from other brands. That's the main drawback I see. I also don't personally love how Apollos sound.

My personal rig is a Quantum and a UAD Satellite Quad, because I wanted to keep my UAD plugins when I sold my Apollos.

I love it and am wanting of nothing. The only thing that could be improved is my PC itself at this point, it's slightly behind what's currently being sold.

There's my two cents anyway, having owned both.

Quantum 2626 TB3. Quite satisfied with mine. No direct monitoring however...

Hello, .

i quote the entire reply because this is exacly my fear and what i think regarding uad ecosystem. ancient dsp processors with embarassed math performance. i think that just with a decent ARM cortex you can have 8 times the power of a single uad duo... UA really know this. it's business.

Anyway... is significant for me when reading an owner that want to SELL their Apollo. After all.... it's not the best of the best out there if people want to sell it and you are not the first that i read... a lot of owners tends to resell it.

Own an Apollo is one of my whim after years of pretty standard hardware or integrated audio... but for what i read on forums... it's not worth the continuous price and the DSP resources is completely out of any reason today. for what i remember it's the same exact chip used in Creamware Pulsar platform, TC Electronics powercore and the very first uad.

I have this specs on my daw.

10900k > 11900k ( xeons are not what xeons have to be nowadays )
z490 aorus xtreme (TB3 integrated onboard and intel certified)
64Gigs ram ( standard xmp 3200 cl14 ).
windows 10 pro 20H2

i own even a threadripper/ryzens but... despite the raw power and core count difference even on 5000 cpus some problems persist vs intels chips.

can you give me any advice for a good cable to connect/convert the Quantum TB2 on this TB3 Board ?. and i really like presonus studio one 5.1 it's between Logic pro , a little bit of cubase and well organized like protools.

Thank you.

Hello, .

thank you... this is already considered. for 500 bucks i don't know how presonus can produce it and earn money. i think i end to buy this one because TB3.

- No direct monitoring;
- Headphone outs is simply a copy of main out (no dedicated headphone chip);
- No metering leds or talkback like the older ones;
- TB3 connector but still TB2 inside;
- You still need a TB3 cable. I didn’t skimped on this and bought a pricey Startech TB3 40gbps 6 ft active cable (even if you can use a lesser 20gbps non-active). Heard some stories of flakey or loss of connections with cheaper cables.

So, less features than the older Quantum, barebones...

I repeat, no monitoring so a DAW must be fired to hear the inputs but the latency is so low that it’s not a problem for me. In fact, I’ve simply created an AUlab config if I don’t need to record (MacOS here). Don’t worry, Windows will be able to use it for system sound (Web, YouTube, etc...)!

Ahh, really ?. have only the type-c of tb3 but this uses a tb2 chipset ?... i suspect that the apollos units uses usb 1.0 at this point because the latency performance seems worst than a usb interface. incredible.

This is very interesting. where you have read about this thing ?.

really thank you for that !.

It's either going to be the Startech or the official Apple one. I have both and one works with some interfaces, the other works with other interfaces.

I'd have to drag the rack into the other room with the TB3 computer so I haven't tested it, therefore I can't give a definite recommendation.

If you contact Presonus directly I bet they would give you some tried information.

Apollo is completely capable of "great sounding records" but so is Quantum, and so on. To me it's more about workflow, value, things like that, features.

Apollo was my first "good" interface and it "was" a gamechanger for me, but I was stepping up from some crusty old tech. They sell a lot of promises. You have to be aware of some of the marketing, and make your own educated choice. Knowing the limits of the SHARC DSP is a good place to start, you can factor that in to your value ideas.

Well it all goes both ways. Ive been watching the Quantum's and Motu's for a bit online as I wanted to try one out. I see a lot of Quantum's being sold as well, and I've asked a few of those people and they're actually trading up for Apollo's... or something completely different. I think Monkey's experience is with gen 1's, maybe 2's? I'm still using Apollo's as they're rock solid, sound great, and are flexible. I mainly track with hardware so it's the only reason I'd considering switching to something simpler... Quantum's might be good, but I've read a lot of mixed things...

Still. 599 for the 2626 is an amazing price. The Apollo DSP is old but does exactly what its supposed too. Also, I buy like 1 or maybe 2 plugins a year with coupons, not expensive at all, just cant go nuts. I don't need to many more plugins but am more into utility stuff at this point. No issues with the latency, 0 latency monitoring is my only concern.

Yeah I have a "regular" Quantum, the first one. It's hard to go completely for a purchase based on online discussions, for many reasons including subjectivity, different levels of ability, computer knowledge, recording styles etc.

The thing I love the most about the Quantum is it's "stripped down." That really works for me. I didn't use the Console plugin area ever, at all, so Apollo was completely wasted on me. I just don't need it. I use Cubase/Quantum in the same way now, but everything is native, and it's all in the same window.

The thing I really liked about Apollo is it seems pretty rock solid. It never burped or farted on me. But the Presonus hasn't, either.

I’ve made searches and also contacted Presonus support about TB3 link speed: the Quantum is still TB2 (or maybe even 1?) inside.

I think that most Thunderbolt interfaces are probably built the same way. 10 gbps is plenty sufficient for a gazillions of in/out request from a computer. So, connector is TB3 (USB-C type) but no way any audio interfaces needs 40 gbps unless I’m mistaken! It’s another game for the video field where TB3 throughput is sometimes needed.

And don’t believe the hype about Thunderbolt equals low latency automatically. It’s up to the driver designers to make it efficient and implement low RTL.

In this regard, the Quantum 2626 delivers: a buffer setting of [email protected] KHz yields 3.4 ms RTL in Windows or MacOS.

For small projects, I even use 32 (or 48, Reaper permits to set the CoreAudio buffer) and there is no clicks and pops.

Better, it comes with Studio One Artist with its hybrid audio engine: low latency for real-time and higher buffer for processing/mixing.

Sure, the king of the hill for driver efficiency is still RME but in its price range, the 2626 is quite capable.

Oh i meant your Apollos were gen 1 or 2... but yea, the OG Quantum is prob still the higher quality out of the 2... I think I'll pick up a 2626 next year to try out. With the 2 sets of adats, I think I ferrofish pulse would pair perfectly... SPDIF for my dbox monitoring, if I decide to hang on to it...

Woo . thank you so much to everyone that replies to me with or without quoting my questions. Thank you.

I just quoted only this reply just to not generate a very long reply with all quotes. but again., thanks to everyone.

Yes, honestly 10gbps ( real and effective bi-directional without overheads ) is overkill for audio even on 384Khz i think... but i know that the thunderbolt protocol ( when is real end full on thunderbolt ) have enormeous advantages in developing drivers to minimize RTL because the way they work compared to USB 3.2 gen2 2x2 ( and for sure even USB4 ) etc etc etc..
thunderbolt 10Gbps is not comparable in anyway to USB 10Gbps ( on the paper ehh ) . i remember the first thunderbolt interfaces that still ( a passthru to firewire ) and I HOPE... that in TB2 and expecially in TB3/4 this ( trick ) is completely and finally gone forever.

Ok guys, i have sufficent information to decide between Apollos and Quantum.

I love RME and i owned the beautiful FF800 for 2 years before to sell it years and years ago... an incredible mistake. i know that have fantastic drivers for both Winos and Macos.

thank you to everyone.

Dan.