The No.1 Website for Pro Audio
 All  This Thread  Reviews  Gear Database  Gear for sale     Latest  Trending
Audio Interface - Low Latency Performance Data Base
Old 30th May 2019
  #3241
Lives for gear
First off. Very well done TAFKAT. Providing all of this information and maintaining this thread for many years. I very much apriciate what you have put together here. Thank you very much!

This thread was a contributing factor to a recent purchase of RME MADI FX, and RME QS8. Lowest lag system with good routing I am aware of.

Now I wanted to take this a step further. What is the lowest lag MIDI? With the AD/DA system I assume it's the one MIDI port on the QS8. But what about an 8 in/out system?

I currently have an AL88 over copperlan and Edirol UM-3EX. Is there anythinb better?

My goal is to get the LAG of electronic drums as low as possible because most classic triggers require about 3ms to decide how loud they are hit. Past 4.5ms total between hitting a drum and hearing the sound makes playing E-drums a drag.

The one positive about this for the sound system is that the output of converters tends to be a faster operation than the input.

An additional twist I hope to fit in that 4.5ms is the ability to take the raw drum sound, run it through outboard, back into the DAW, then to the monitoring system. Getting that bag of tricks under 4.5 ms is near impossible.

I have a few drum brains, Some are 5 Pin only, Some also have USB. I find that comming out of the brains with 5Pin to be better than the built in USB.

I would like to hear your thoughts on the subject.
Note: BFD3/Cubase are my software of choice. But am open to suggestions if they help out the E-drum lag.
Old 30th May 2019
  #3242
Quote:
Originally Posted by elegentdrum View Post
First off. Very well done TAFKAT. Providing all of this information and maintaining this thread for many years. I very much apriciate what you have put together here. Thank you very much!

This thread was a contributing factor to a recent purchase of RME MADI FX, and RME QS8. Lowest lag system with good routing I am aware of.

Now I wanted to take this a step further. What is the lowest lag MIDI? With the AD/DA system I assume it's the one MIDI port on the QS8. But what about an 8 in/out system?

I currently have an AL88 over copperlan and Edirol UM-3EX. Is there anythinb better?

My goal is to get the LAG of electronic drums as low as possible because most classic triggers require about 3ms to decide how loud they are hit. Past 4.5ms total between hitting a drum and hearing the sound makes playing E-drums a drag.

The one positive about this for the sound system is that the output of converters tends to be a faster operation than the input.

An additional twist I hope to fit in that 4.5ms is the ability to take the raw drum sound, run it through outboard, back into the DAW, then to the monitoring system. Getting that bag of tricks under 4.5 ms is near impossible.

I have a few drum brains, Some are 5 Pin only, Some also have USB. I find that comming out of the brains with 5Pin to be better than the built in USB.

I would like to hear your thoughts on the subject.
Note: BFD3/Cubase are my software of choice. But am open to suggestions if they help out the E-drum lag.

I suggest creating a new thread. It's off topic here. Along with all the other posts about Thunderbolt connectors etc. This is supposed to be about audio interface RTL. That's it.
Old 30th May 2019
  #3243
Lives for gear
 

Quote:
Originally Posted by TAFKAT View Post
Just to finish off, TB3 using a USB-C interconnect port doesn't necessarily mean its being pushed through a USB-C controller, 3rd party or otherwise, its simply the port connector they chose to use for wider acceptance, which has resulted in massive confusion I might add.
Yes, both standards can be pushed through the USB-C port round the rear and I was aware that some earlier boards were doing so in a combined solution that was causing me headaches.

Quote:
Originally Posted by TAFKAT View Post
I can't see how/why the TB throughput from the PCIe cards are being pushed through a USB-C Controller instead of the dedicated TB controller on the card, makes absolutely no sense to me.
I never stated that, I said onboard solutions, add in cards are not onboard solutions, a fact you've seen me rant about before I imagine. I agree, why would a drop in card be routed via another route.

That said I take your point, I've not used that board you mentioned and it does explicitly state that it's running an Intel connector and nothing else is being routed through it so, yay. It's a good one.

Still, sooner it's native the sooner all this confusion becomes irrelevant. I'm surprised still that they just didn't make it standard on the "enthusiast" platform on the X299 refresh and keep it optional lower down. If they are going to keep treating it like a niche product for enthusiasts, they might as well get behind it as one.

Quote:
Originally Posted by Dallon426 View Post
I suggest creating a new thread. It's off topic here. Along with all the other posts about Thunderbolt connectors etc. This is supposed to be about audio interface RTL. That's it.
... ok, point taken. I now realise this isn't the "today we build our PC thread"... I shall STFU again.
Old 31st May 2019
  #3244
Gear Maniac
 

Quote:
Originally Posted by TAFKAT View Post
A direct link to the discussion would be of value for those interested , but we are going way off tangent. TB implementations on some laptops has always been a minefield, and its directly related to how the Displayport interconnect is implemented, which is completely irrelevant to the discussion here, or my results I presented.

This is getting exhausting,
Sorry about that @ TAFKAT .

My point, which was not clearly written, was that laptop GAMER issues with TB are principally around: bandwidth and "compatibility", not latency.

It seems TB AUDIO users probably contend with "compatibility" but not bandwidth or latency. I sense general TB situation is improving and moving TB from controller board to the CPU should resolve more "compatibility" issues, both for laptops and desktops.

Back to our scheduled programming.
Old 31st May 2019
  #3245
Lives for gear
 
TAFKAT's Avatar
 

Quote:
Originally Posted by vanallen View Post
Sorry about that @ TAFKAT .

My point, which was not clearly written, was that laptop GAMER issues with TB are principally around: bandwidth and "compatibility", not latency.
All good V, I wasn't referring to you re the discussion getting exhausting , you got caught in the cross fire :-)

TB issues for audio are the same as any other device, implementation / compatibility. I personally have not experienced or consider latency being an issue , which was part of the frustration. I have had numerous curves with getting its settled and working tho.

As already noted by a few others, the sooner it becomes native the better, but I wouldn't be holding my breath, and with the wider adoption on non Intel platforms now, I think we will still be navigating some curves for a while yet.

I digress

Quote:
Back to our scheduled programming.
Indeed, and I will have another TB interface on the bench this week to test :-)

Old 1st June 2019
  #3246
Quote:
Originally Posted by TAFKAT View Post
All good V, I wasn't referring to you re the discussion getting exhausting , you got caught in the cross fire :-)

TB issues for audio are the same as any other device, implementation / compatibility. I personally have not experienced or consider latency being an issue , which was part of the frustration. I have had numerous curves with getting its settled and working tho.

As already noted by a few others, the sooner it becomes native the better, but I wouldn't be holding my breath, and with the wider adoption on non Intel platforms now, I think we will still be navigating some curves for a while yet.

I digress



Indeed, and I will have another TB interface on the bench this week to test :-)

I think we will be seeing USB 4 being used more than Thunderbolt

https://www.youtube.com/watch?v=Nr3FbED_vwg
Old 2nd June 2019
  #3247
Lives for gear
 
Jim Rosebrook's Avatar
 

Quote:
Originally Posted by Dallon426 View Post
I think we will be seeing USB 4 being used more than Thunderbolt

https://www.youtube.com/watch?v=Nr3FbED_vwg
And it's possible that USB4 and Thunderbolt 3 will essentially converge.
Old 3rd June 2019
  #3248
Lives for gear
 
TAFKAT's Avatar
 

Quote:
Originally Posted by Dallon426 View Post
I think we will be seeing USB 4 being used more than Thunderbolt
We are yet to see a USB-C interface , those on the market using a USB-C connector are still actually using a USB2 controller , so I can't see any urgency towards navigating to USB 4.

The connector will still be the USB-C connector , throwing even more confusion regards compatible cabling into the mix.

Old 3rd June 2019
  #3249
Gear Maniac
 

here's some xair 18 latency measurements, though it has dsp so maybe not the most important measurements.

Probably never going back to computer based latency ever again. I underestimated just how much of a difference zero latency can be. I feel so much more connected with what I'm playing and everything I track is more in time. xair is also proving to be a super handy piece of kit like hot damn. I liked the 18 so much I bought the 12 channel model just to keep in the living room for movies (compression and EQ). I know UAD is probably better in the pre and DA sections but this is more than enough for me. Feels like a solution to all my problems I've been having with latency. The only thing I don't care for is that, and it may be placebo, but I don't feel it sounds as good as my umc404. Could be placebo but there definitely seems to be a sort of congestedness to the sound.

Latency is the mind killer.
Attached Thumbnails
Audio Interface - Low Latency Performance Data Base-xair.png  

Last edited by FoundationsAudio; 3rd June 2019 at 02:52 PM..
Old 4th June 2019
  #3250
Gear Head
 
broken head's Avatar
 

Not to throw Kerosene on a fire, but I concur that from our experience, all else being equal, Thunderbolt and PCIe are rough equivalents. An AES16e + Aurora, or AuroraTB should be near identical in latency performance. In fact, for all intents and purposes, the LTTB card is a host to the PCIe core used in the AES16e. Thus they use the same firmware and drivers.
I can only suspect that when I made a reference to "properly implemented thunderbolt" (and if I used that language, my apologies, it is clumsy and vague) I would have been referring to the potential to utilize firewire or USB cores within TB implementation, which has potential to port some of the problems innate to those connection options over. The beauty of thunderbolt is that it can be a single wire, external format that delivers PCIe performance and sub-kernel transfers. .
Anyway, if I obfuscated anyone's understanding through clumsy word selection, my apologies...
Oh, one more thing, latency performance with the Aurora(N) would be no different than the original Aurora, FWIW
Old 5th June 2019
  #3251
Lives for gear
 
TAFKAT's Avatar
 

Quote:
Originally Posted by broken head View Post
Not to throw Kerosene on a fire, but I concur that from our experience, all else being equal, Thunderbolt and PCIe are rough equivalents. An AES16e + Aurora, or AuroraTB should be near identical in latency performance.
Ha, more like throwing a fire blanket over it :-)

No argument that TB and PCie are roughly equivalent, which is exactly what my testing showed, 0.09ms measure RTL between the 2 Lynx configurations. There was a slight variable with the benchmark results as noted, but close enough to cooee.


Quote:
In fact, for all intents and purposes, the LTTB card is a host to the PCIe core used in the AES16e. Thus they use the same firmware and drivers.


Quote:
I can only suspect that when I made a reference to "properly implemented thunderbolt" (and if I used that language, my apologies, it is clumsy and vague) I would have been referring to the potential to utilize firewire or USB cores within TB implementation, which has potential to port some of the problems innate to those connection options over.
Thank you for the clarification.

Quote:
Oh, one more thing, latency performance with the Aurora(N) would be no different than the original Aurora, FWIW
Noted.

Thanks Paul.


Last edited by TAFKAT; 5th June 2019 at 12:31 AM..
Old 5th June 2019
  #3252
Lives for gear
 

@ broken head Thank you for the explanation!. Appreciated.

All the best,

Michal
Old 6th June 2019
  #3253
Lives for gear
 
zephonic's Avatar
Vin, I hadn't looked here for a while, and see that the UAD Apollo Twin is doing much better than previous-gen Apollo's, that's a nice surprise.

When did you test it? I'm trying to find your report on it, but knowing when it was tested would make that search a bit easier.
Old 7th June 2019
  #3254
Lives for gear
 
TAFKAT's Avatar
 

Quote:
Originally Posted by zephonic View Post
Vin, I hadn't looked here for a while, and see that the UAD Apollo Twin is doing much better than previous-gen Apollo's, that's a nice surprise.

When did you test it? I'm trying to find your report on it, but knowing when it was tested would make that search a bit easier.
Hey Z,

September 2017 is when I tested the TB Apollo Twin : Here

I have a new X8 here to run some numbers on over the next week.

Old 7th June 2019
  #3255
Lives for gear
 
zephonic's Avatar
Thanks, Vin!

I always liked how the Apollos sounded, but as I rely heavily on VI’s (and thus LLP) I felt they were not a good fit for me. If that issue has been resolved, it may be time to have another look at it.

Waiting to hear you report back on the x8!
Old 8th June 2019
  #3256
Gear Maniac
 
Secret80sMan's Avatar
 

Quote:
Originally Posted by TAFKAT View Post
Hey Z,

September 2017 is when I tested the TB Apollo Twin : Here

I have a new X8 here to run some numbers on over the next week.

Extremely curious to see your Apollo X8 results. Looking forward to not just the numbers but your always thoughtful insight Vin.
Old 12th June 2019
  #3257
Quote:
Originally Posted by TAFKAT View Post
Hey Z,

September 2017 is when I tested the TB Apollo Twin : Here

I have a new X8 here to run some numbers on over the next week.

Hey Vin, you never got around to testing an Arrow, did you?
Old 4 weeks ago
  #3258
Here for the gear
 

Very helpfull stuff! thx for that!

I checked out your dawbench page and was a bit shocked at the performance of winxp sp3 vs the others. I knew it performed much better audio wise then windows 7, i could clearly see the difference when i switched to 7 buti would never expected it to be this much better then all of them including even osx.
quite shocking tbh..
Just a while ago i was debating on
making an xp installation just to be able to run 32 bit audio apps on there.
So that def gonna happen now
Old 4 weeks ago
  #3259
Lives for gear
 
TAFKAT's Avatar
 

Hey All,

O.K, finally got the UAD X8 numbers finalised.



Charts above in comparison to the Apollo Twin, not a lot of difference in overall performance, but there are some interesting points I will cover in detail.

Most obvious will be the rating has dropped , even though the actual benchmark results are a little up compared to the earlier driver. That is due to the higher measured RTL, despite the reported I/O and RTL being identical to the previous driver used on the Apollo Twin.

That brings up a few points, first being that the reported and measured results on the Twin were identical, meaning that the reported I/O and RTL was identical to the sample to the measured.

In this instance there is a consistent additional .79ms in the measured RTL , not huge but no longer reported. That could be the AD/DA is not being reported on the X8 by the driver , or some additional safety buffers. Difficult to determine where exactly the additional latency is assigned as the reported I/O values to the DAW remain identical to the Twin.

This being the first TB3 interface I tested on my reference system was my immediate hurdle as I have a unicorn native TB2 card on my system , so all previous TB interfaces were tested natively on a TB2 card. I have successfully configured quite a few TB2 interfaces on TB3 controller cards for clients systems , but this was going to be the complete opposite with a TB3 interface needing to be scaled back to run on a TB2 controller card.

In theory the TB3 to TB2 adapters are bidirectional, and I use the well known and officially recommended Startech adapter on Windows 10, and I also have an Apple adapter here for cross reference. The Startech has worked flawlessly on all previous installations where I was going from a TB3 controller to a TB2 interface , so I thought that would be my first cab off the rank.

The X8 was completely ignored by the system , which wasn't a good start , but luckily the Apple adapter worked perfectly, go figure. This is in complete opposition to all previous TB2 interfaces being adapted the other way.

With that out of the way I was able to quickly measure the RTL using the RTL Utility and compare the numbers to the numbers I had achieved on the X8 in a native TB3 to TB3 configuration ( on the clients new system ).

The difference across the board on all respective measured latency settings, 1 sample, yep, 1 additional sample using the TB3 to TB2 adapter. I think that puts all of the hysteria of supposed correct TB implementation to bed, once and for all if it hadn't previously !!

IMO UAD are very particular and overly stringent in their supported TB qualified configurations. Its obvious they have a very narrow internal R&D.

On the Apollo Twin for example, they only support the unit on Windows 10 attached to a TB3 controller card via a Startech Adapter to the TB2 unit , they do not support TB2 natively. On the X8 its the opposite , X8 is only supported via a native TB3 Controller. They also specify what versions of Windows 10 , which to some degree I can understand because W10 is a moving target, but its still overly narrow, IMO.

Now I can understand that to a degree, but some lock systems down and don't immediately hop onto the latest and greatest W10 iteration, so in short if you have any issue with the unit connectivity/performance, their only advise will be to upgrade Windows before they will support the interface. Well, I could write a chapter how that isn't necessarily the wisest move, but I digress.

So in summary, very similar performance to the previously tested Apollo Twin TB2 , I/O and RTL has crept up a touch , connectivity on Windows, as with anything TB, could be a crap shoot.

Old 4 weeks ago
  #3260
Lives for gear
 
zephonic's Avatar
Quote:
Originally Posted by TAFKAT View Post
O.K, finally got the UAD X8 numbers finalised.
Thanks, Vin!

Sounds like the Apollo series has become a viable option for VI users.

So the slightly reduced score is due to the higher RTL, even though voice count has gone up a bit?
Old 4 weeks ago
  #3261
Gear Maniac
 
Secret80sMan's Avatar
 

Thanks for the lowdown on the X8 Vin. Your efforts are always appreciated.
Old 4 weeks ago
  #3262
Here for the gear
Hi all,

On the hunt for an interface and realized a 2i2 is not going to cut it any longer.

Looking through all the charts, seems like the two best options in my budget would be the RME Babyface Pro, or the Antelope Discrete 4.

I don't see anything clocked for the Discreet 4, but the Discreet 8 is near the top of the list.

Since I'm financing anyway, would it be worth it to go the Antelope route?

I would be using a TB2 to TB3/USB-C adapter cable to plug into the USBC port on my playback machine.

I've got a limited timeline to go on the Antelope at a $1200 price point ($1300 with taxes) vs the $1050 Babyface Pro.

Any ideas guys and gals?

Thanks,

KZ
Old 4 weeks ago
  #3263
Lives for gear
When you read Antelope threads there are a lot of complaints about interfaces being abandoned in favor of constantly expanding the lineup and stability issues. RME has a much smaller lineup and supports interfaces seemingly forever, and are known for stability and support. So with that in mind, I know which company I'd choose.
Old 4 weeks ago
  #3264
Gear Guru
 
monkeyxx's Avatar
Antelope might bite you in the a$$ and that would be a painful mistake. Some people get lucky. I wouldn't go there. I got burned by an earlier company they're related to called Aardvark, lost a lot of money in that exchange.
Old 3 weeks ago
  #3265
Lives for gear
Quote:
Originally Posted by monkeyxx View Post
Antelope might bite you in the a$$ and that would be a painful mistake. Some people get lucky. I wouldn't go there. I got burned by an earlier company they're related to called Aardvark, lost a lot of money in that exchange.
I make it a rule never to buy interfaces named after animals.
Old 3 weeks ago
  #3266
Gear Head
 

Quote:
Originally Posted by studer58 View Post
I make it a rule never to buy interfaces named after animals.
Damn, I've been dreaming about a Crane Song... Back to the drawing board.
Old 3 weeks ago
  #3267
Quote:
Originally Posted by camacozie View Post
Damn, I've been dreaming about a Crane Song... Back to the drawing board.
It's not an interface.
Old 3 weeks ago
  #3268
Gear Head
 

Quote:
Originally Posted by DAW PLUS View Post
It's not an interface.
Technically, you're right; it's a converter. But still, she's a beaut.

http://www.cranesong.com/Hedd_Quantum.html
Old 2 weeks ago
  #3269
Lives for gear
 
Gomjab's Avatar
 

I would love to see results for a MOTU 828es or 8pre-es
Old 2 weeks ago
  #3270
I've tested a few motus and I wasn't impressed. Search through my posts in this thread and you'll find em.
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump