Topic: RME Asio stability/latency vs. UAD Apollo vs Orion32

Hello, I owned a few Firefaces and a 9632 back in the day and loved them. I'm currently running an Apollo 8 fed by ADAT from an Antelope Orion32. It's somewhat stable as long as I don't chain too many plugins in a row, but I'd like to get the maximum amount of power out of my system. I'm considering getting an RME PCIe card for better stability/latency. I could feed it with audio by ADAT (or MADI) from the Apollo and Orion - I still want to use the Apollo for its near-zero UAD effects for monitoring. I'm on Windows 10. i9 x7940, Cubase 9.5, Asus x299 PCIe mobo, mostly large cpu-intensive sessions.

Could anyone share how the RME PCIe driver latency/stability compares nowadays to the Apollo and/or Orion?

Are the newer PCIe cards far superior to the drivers on the old PCI cards? It seems the two options are the MADI (I think too expensive for me) or the Raydat. If anyone has any insight, I'd appreciate it.

2 (edited by ramses 2018-10-22 09:04:07)

Re: RME Asio stability/latency vs. UAD Apollo vs Orion32

garbage914 wrote:

Hello, I owned a few Firefaces and a 9632 back in the day and loved them. I'm currently running an Apollo 8 fed by ADAT from an Antelope Orion32. It's somewhat stable as long as I don't chain too many plugins in a row, but I'd like to get the maximum amount of power out of my system. I'm considering getting an RME PCIe card for better stability/latency. I could feed it with audio by ADAT (or MADI) from the Apollo and Orion - I still want to use the Apollo for its near-zero UAD effects for monitoring. I'm on Windows 10. i9 x7940, Cubase 9.5, Asus x299 PCIe mobo, mostly large cpu-intensive sessions.

Could anyone share how the RME PCIe driver latency/stability compares nowadays to the Apollo and/or Orion?

Are the newer PCIe cards far superior to the drivers on the old PCI cards? It seems the two options are the MADI (I think too expensive for me) or the Raydat. If anyone has any insight, I'd appreciate it.

1st of all .. all RME drivers have excellent stability and very low round trip times.
USB and Firewire are on par with PCIe based solution.
You can have a look at my excel below which RTTs the ASIO driver reports to the DAW at 44.1 kHz.
With higher sample rates the RTT times are even lower.

If you use i.e. recording interfaces that are either pure digital like the RayDAT or digital to 99% like the HDSPe MADI FX,
then you might need to add the values for A/D and D/A conversion to get the full latency.

If you are i.e. using a mic and want to calculate the latency "mouth to ear" then you need to add the values of A/D and D/A conversion of your mic preamp and/or AD/DA converter additionally to the RTT time that the ASIO driver reports to the DAW.

For this reason I added 2 more colums to my excel for 2 of such examples.

So it depends a little on your setup.

For a first impression I recommend you to look first only at the RTT times reported by the ASIO driver.

https://www.tonstudio-forum.de/index.php/Attachment/2343-UFX-UFX-RayDAT-Latencies-v2-jpg/

One of the big advantages of the HDSPe MADI FX is, see product description on web, that
- it uses only system resources for audio channels that are active. This makes the card very attractive to me.
- and its the only PCI/PCIe card which has a full FX chip on board

But you can use also the RayDAT if you mainly want to connect devices via ADAT.

The MADI card is more expensive on the 1st glimpse, true, but on the other hand it might pay out on the long run.
But then you also need to consider, that the other devices also need to support MADI.

To replace the whole audio infrastructure might be too expensive.
Therefore you could i.e. consider to get the RayDAT and then connect other devices by ADAT.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

Re: RME Asio stability/latency vs. UAD Apollo vs Orion32

Thank you @ramses. That's very helpful.

Do you know if there is any latency info on the USB Madiface compared to the HDSPe Madi or the RayDat? And also, any info on latencies at 1024m 2048, or even 4096? I mostly use larger buffers.

That feature of the HDSPe Madi to only use resources on active channels is very appealing. Do you think the USM Madiface would use many resources anyway? It seems like a very simple box.

Also,do you have any guess on if I'd see an improvement over my Orion or Apollo drivers?

4 (edited by ramses 2018-10-23 09:32:52)

Re: RME Asio stability/latency vs. UAD Apollo vs Orion32

EDITED: pls re-read.

Do you use such high ASIO buffers at 44.1 kHz sample rate ?

If yes: then you do not need to worry much about RTT latency. Then you want simply high ASIO buffersizes for i.e. important recordings, where safety and reliability counts. High ASIO buffersizes ensure least CPU load and most stability, shall your system start a background job or other tasks concurrent to recording.

The Excel does not contain higher ASIO buffer sizes because I want to focus on those values which are especially interesting for people who need low RTT latencies of under 10ms. This is i.e. needed if you work with virtual instruments (VSTi's) like virtual guitar amps, drums, keyboards, etc. You can expect that the RTT times for higher ASIO buffer sizes are also very good, when the RTT values are excellent with smaller ASIO buffer sizes.

You do not need to be worried about system resource consumption. Every modern (not historic) PC has enough CPU power to drive a recording interface, no matter whether its Firewire, USB2/3. I only wanted to highlight this very unique feature of this card.

For me personally the HDSPe MADI FX is interesting, because I try to fine tune and optimize everything whats possible.
I am always working with a few VSTi's in my recording projects during mixing and mastering.

I was using two UFX+ and an ADI-2 Pro before, connected to my PC (one UFX+ for recording, the other one for mobile use in my guitar rack). I had no issues with that big amount of channels connecting all to this PC.
And if you see, that even projects with a large amount of tracks and VSTs (2 Cubase VSTs per track) are able to run with the UFX+ at about the same CPU load as with a RayDAT PCIe based card, then you can see that the USB drivers are in very good shape.
https://www.tonstudio-forum.de/blog/ind … cks-de-en/

So you should IMHO 1st concentrate to pick a recording interface that serves your demand best.
Some people prefer USB some people prefer PCIe based cards, depending on their demands.

I personally don't know the MADIface USB, but it surely will also be a very nice product. I think the beauty of this recording interface is, that you simply have a TotalMix FX capable interface with only very few demands in terms of power consumption. It doesn't need to drive any headphones or does not need to provide phantom power to Mics. Looks like a straight possibility to enter MADI world with a small card that you can connect to Laptops or PCs .. powered by USB2 bus.

I can not judge about Orion/Apollo drivers, but keep in mind with RME you get
- TotalMix FX together with the drivers, TM FX is extremely well
- The new TotalMix Remote
- Digicheck tools
- with Digicheck you also get Global Record (for Windows at least) which is a light weight recording application for very reliable recording (something like Durec but on the PC)
- with MADI the ability to remote control preamps and converters without requiring additional MIDI cabling
- with the Octamic XTC nearly full integration of the device into the TM FX instance of your RME recording interface (see AUX device)
- for all other RME devices RME MIDI remote
- if you do not use MADI, then you can remote control devices via MIDI

In total: a very round and complete package and with HW/SW being maintained excellent for a very long time.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

Re: RME Asio stability/latency vs. UAD Apollo vs Orion32

Thank you again for your generous response. I do use high buffers - usually 2048. I'm more hoping that if the RME drivers are more efficient and stable, I'll be able to get more plugin processing happening before maxing out my system. I mainly max out when I have a lot of plugins in a chain - hoping the RME driver might help a little with that relative to the Apollo or Orion drivers. When I record VSTs, I turn on Low-Latency monitoring in cubase, which works pretty well.

I wouldn't mind getting an RME card that has midi, as I know RME midi times very accurately. But I instead decided to go with the Madiface. Just ordered it and looking forward to seeing how it does.

A bit off topic, but do you have a recommendation for a low-cost midi interface that is rock solid?

Re: RME Asio stability/latency vs. UAD Apollo vs Orion32

garbage914 wrote:

A bit off topic, but do you have a recommendation for a low-cost midi interface that is rock solid?

Sorry, no idea, maybe somebody other has a good recommendation.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

Re: RME Asio stability/latency vs. UAD Apollo vs Orion32

Which Madiface? The USB or the XT?
You can send Midi over MADI and use the Midi connection on any other Madi Device.
The XT has Midi connections afaik.
Georg