Topic: Thunderbolt USB

Would FFUC be able to use Thunderbolt and will it make a difference?

Re: Thunderbolt USB

No and no.


Regards
Daniel Fuchs
RME

Regards
Daniel Fuchs
RME

Re: Thunderbolt USB

Will FF400 be able to us Thunderbolt? I'm trying to decide between an FF400 and UC.

Re: Thunderbolt USB

The FF400 is a Firewire device only, not Thunderbolt...
TB won't make any such device "faster"

Regards
Daniel Fuchs
RME

Regards
Daniel Fuchs
RME

Re: Thunderbolt USB

Thank you, Daniel.

Re: Thunderbolt USB

I thought thunderbolt is a much faster connection which would allow lower latency and better overall performance of DAW's.
Maybe a tech can point out what thunderbolt will improve and what RME's attitude is towards tb. Will there be a fire face tb any time soon?

Re: Thunderbolt USB

TB has wider bandwith/data rates. Audio devices don´t need that amount of data, so no difference, put simply.

Vincent, Amsterdam
https://soundcloud.com/thesecretworld
Babyface pro fs, HDSP9652+ADI-8AE, HDSP9632

Re: Thunderbolt USB

I thought thunderbolt is a much faster connection which would allow lower latency and better overall performance of DAW's.

This is a myth. Thunderbolt canot be faster as PCI express interfaces. RME's USB & FireWire interfaces are already close to the unique low latencies of RME PCI express cards.

Maybe a tech can point out what thunderbolt will improve

The costs. We are speaking of several hundred USD for a comparable Thunderbolt-equipped system. Plus the cost of the Thundebolt cable - which is limited to 2 m at the moment.

If you want a real Thunderbolt interface now, buy the new external Sonnet PCI express chassis and use a RME HDSPe AIO/RayDAT/AES/MADI with it. We've demonstrated such a system with 390 channels on the new HDPSPe MADI in Frankfurt and it worked great.

http://www.sonnettech.com/product/images/echochassis_overview.png

But it's a Mac only technology in the moment. It works with a Windows OS just like PCIe, this means it lacks hot swapping. The system has to boot or reboot with the connected device. Plug and play - like on FireWire or USB - is not supported with Thunderbolt.




best regards
Knut

Re: Thunderbolt USB

"Close to" is not the same as "equal to". I, and many other composers,  am running huge templates with demanding virtual instruments/sample libraries and I need to use the lowest buffer size I can, usually 128 but sometimes 256 with my HDSPe-AIO.

I have my HDSPe-AIO in aSonnet chassis btw and I CAN hot swap and "plug and play".

I know that at its heart RME is PC-centric but if RME does not design a TB interface (Lightpeak on PC, right?) I think it is mistake.

Composer, Logic Certified Trainer Level 2
author of "Going Pro With Logic Pro 9"
www.jayasher.com

Re: Thunderbolt USB

Interface bandwidth (not "speed") does not directly translate into lower latency. A TB equipped FF would not automatically enable your system to run lower latencies.

Regards
Daniel Fuchs
RME

Regards
Daniel Fuchs
RME

Re: Thunderbolt USB

RME Support wrote:

Interface bandwidth (not "speed") does not directly translate into lower latency. A TB equipped FF would not automatically enable your system to run lower latencies.

Regards
Daniel Fuchs
RME

Please check your email when you have a minute.

Composer, Logic Certified Trainer Level 2
author of "Going Pro With Logic Pro 9"
www.jayasher.com

12

Re: Thunderbolt USB

> I have my HDSPe-AIO in aSonnet chassis btw and I CAN hot swap and "plug and play".

Knut clearly said 'under Windows'. Of course that works under Mac.

Regards
Matthias Carstens
RME

Re: Thunderbolt USB

MC wrote:

> I have my HDSPe-AIO in aSonnet chassis btw and I CAN hot swap and "plug and play".

Knut clearly said 'under Windows'. Of course that works under Mac.

Thanks for the clarification, I missed it.

Composer, Logic Certified Trainer Level 2
author of "Going Pro With Logic Pro 9"
www.jayasher.com

Re: Thunderbolt USB

BTW, UA says:

"Thunderbolt provides lower latency, reduced audio buffer size, improved performance, and greater UAD plug-in instances versus FireWire. And because Thunderbolt offers many times the bandwidth of FireWire, it allows music producers to connect numerous devices in series with the Apollo interface — including hard drives, processors, and additional computer monitors — all with fast, flawless performance."

So somebody is wrong:)

Composer, Logic Certified Trainer Level 2
author of "Going Pro With Logic Pro 9"
www.jayasher.com

Re: Thunderbolt USB

BTW, that is not "UA says", but "UA marketing says". tongue

Re: Thunderbolt USB

And I think the UA comment also has a lot to do with the extra bandwidth and buffer required to run the Apollo plugins, not as much using it as a pure audio interface.

Leon Herbers
www.dawplus.com

Re: Thunderbolt USB

The difference between Thunderbolt and USB2 is comparable to the difference to PCI express and USB2. This was discussed and measured a million times in the last years. Nothing new or special.

But the initial question above was a different one: "Would FFUC be able to use Thunderbolt and will it make a difference?"
And the correct answer was: No and no. With a Thunderbolt > USB/FireWire adapter it will be still a USB/FireWire interface.

best regards
Knut

Re: Thunderbolt USB

Admin Knut wrote:

The difference between Thunderbolt and USB2 is comparable to the difference to PCI express and USB2. This was discussed and measured a million times in the last years. Nothing new or special.

But the initial question above was a different one: "Would FFUC be able to use Thunderbolt and will it make a difference?"
And the correct answer was: No and no. With a Thunderbolt > USB/FireWire adapter it will be still a USB/FireWire interface.

best regards
Knut

Understood. So we need you to make a Fireface TB. Tack on $500 to the cost and get it out. No hurry, 1/1/13 is fine:)

Composer, Logic Certified Trainer Level 2
author of "Going Pro With Logic Pro 9"
www.jayasher.com

Re: Thunderbolt USB

TB won't surpass any of the PCIe HDSPe interfaces already available (and without the $500 "TB Tax" and $75 cable with a 2Meter limitation).  Granted - it would make a nice addition for a Laptop user, but if you really need TB's bandwidth capabilities for audio, then you're probably using a serious rack-mounted rig and not a consumer laptop with compromised chipsets/CPU's (compromised for power draw vs performance comapred to a proper desktop) for such critical tasks.  You seem to understand that TB uses the PCIe protocol - so it can't be any "Faster" than PCIe - if anything the bridge chips and numerous repeater-chips in the cable itself add some latency compared to PCIe - but not likely enough to affect audio latency.

Otherwise, the UCX via USB to a laptop will likely be within 0.5ms-1.0ms "round trip" compared to a RME TB interface that would cost about 2x as much.  An RME TB interface would obviously be more "Mac Centric" as PC's are slow to adopt the expensive TB chip (further reducing the pool of potential buyers and cranking up the cost even higher for RME to recoupe R&D costs).

The TB-to-PCIe chassis loaded with whatever RME HDSPe cards (and whatever other PCIe cards your heart desires) would seem to make more sense from RME's standpoint - and from an end-user's standpoint (mix and match, etc).  And that's something you can buy TODAY IIRC wink  Go for it! :-)

cool

MADIface-XT+ARC / 3x HDSP MADI / ADI648
2x SSL Alphalink MADI AX
2x Multiface / 2x Digiface /2x ADI8

Re: Thunderbolt USB

Randyman... wrote:

TB won't surpass any of the PCIe HDSPe interfaces already available (and without the $500 "TB Tax" and $75 cable with a 2Meter limitation).  Granted - it would make a nice addition for a Laptop user, but if you really need TB's bandwidth capabilities for audio, then you're probably using a serious rack-mounted rig and not a consumer laptop with compromised chipsets/CPU's (compromised for power draw vs performance comapred to a proper desktop) for such critical tasks.  You seem to understand that TB uses the PCIe protocol - so it can't be any "Faster" than PCIe - if anything the bridge chips and numerous repeater-chips in the cable itself add some latency compared to PCIe - but not likely enough to affect audio latency.

Otherwise, the UCX via USB to a laptop will likely be within 0.5ms-1.0ms "round trip" compared to a RME TB interface that would cost about 2x as much.  An RME TB interface would obviously be more "Mac Centric" as PC's are slow to adopt the expensive TB chip (further reducing the pool of potential buyers and cranking up the cost even higher for RME to recoupe R&D costs).

The TB-to-PCIe chassis loaded with whatever RME HDSPe cards (and whatever other PCIe cards your heart desires) would seem to make more sense from RME's standpoint - and from an end-user's standpoint (mix and match, etc).  And that's something you can buy TODAY IIRC wink  Go for it! :-)

cool

True but TB brings PCI-e level performance to iMac, Mac Mini, laptop users etc. who are not able to use PCI-e. As I say, my HDSPe-AIO (and a UAD card) is in the Sonnet Express Pro chassis and it is working fine, but the chassis was not cheap and the fans are noisy. My tech guy is going to replace them with quieter fans, an additional expense.

Composer, Logic Certified Trainer Level 2
author of "Going Pro With Logic Pro 9"
www.jayasher.com

Re: Thunderbolt USB

The Sonnet Express Pro chassis is like $799 - so it's only $300 more than what you're suggesting RME offer for $500 (and with less overall flexibility compared to an "open" PCIe chassis, and with a very small market of potential buyers).  And as you say, the Sonnet Chassis also opens up these limited "PCIe-less" devices to the PCIe standard as a bonus (not just an RME interface over TB, but any PCIe card you might need access to).  I'd say that benefit alone is worth the extra $300 for the Sonnet PCIe chassis.

iMac, MacMini, and Macbooks are all based around "compromised" laptop chipsets (they favor low-power draw over raw performance).  If you really need the INSANE bandwidth avaiable over PCIe (or TB) for Audio duties, you are likely better off with a desktop and its "full sized" chipsets and faster CPU's IMNSHO.  Otherwise, the USB UCX is (again) likely within 1ms of a TB system (round trip) for any of the "PCIe-less" Apple devices you mention.  Would you pay $500 (likely much more) for 1ms improvement in latency (USB UCX vs an RME TB Interface)?  That's a pretty small niche market IMO - and I can understand RME's reluctance on making a dedicated TB interface for the time being.

Maybe RME will re-consider this in a few more years once TB has gone through a few generation upgrades and has a larger install-base (as in: widespread PC adoption) and lower upfront cost (chipset and cable costs can come down quite a bit IMO).  In the meanwhile, I'll be more than happy to get my hands on the HDSPe-MADI-FX in my PCIe equipped desktop (upgrading from the HDSPe MADI) wink  I couldn't imagine the logistics of dealing with 64 tracks of MADI on a 13" Laptop, much less 3x that with the MADI-FX!  Otherwise, the USB UCX is likely ample for Laptop (and iMac systems based on compact Laptop components) applications...

cool

MADIface-XT+ARC / 3x HDSP MADI / ADI648
2x SSL Alphalink MADI AX
2x Multiface / 2x Digiface /2x ADI8

Re: Thunderbolt USB

Randyman... wrote:

The Sonnet Express Pro chassis is like $799 - so it's only $300 more than what you're suggesting RME offer for $500 (and with less overall flexibility compared to an "open" PCIe chassis, and with a very small market of potential buyers).  And as you say, the Sonnet Chassis also opens up these limited "PCIe-less" devices to the PCIe standard as a bonus (not just an RME interface over TB, but any PCIe card you might need access to).  I'd say that benefit alone is worth the extra $300 for the Sonnet PCIe chassis.

iMac, MacMini, and Macbooks are all based around "compromised" laptop chipsets (they favor low-power draw over raw performance).  If you really need the INSANE bandwidth avaiable over PCIe (or TB) for Audio duties, you are likely better off with a desktop and its "full sized" chipsets and faster CPU's IMNSHO.  Otherwise, the USB UCX is (again) likely within 1ms of a TB system (round trip) for any of the "PCIe-less" Apple devices you mention.  Would you pay $500 (likely much more) for 1ms improvement in latency (USB UCX vs an RME TB Interface)?  That's a pretty small niche market IMO - and I can understand RME's reluctance on making a dedicated TB interface for the time being.

Maybe RME will re-consider this in a few more years once TB has gone through a few generation upgrades and has a larger install-base (as in: widespread PC adoption) and lower upfront cost (chipset and cable costs can come down quite a bit IMO).  In the meanwhile, I'll be more than happy to get my hands on the HDSPe-MADI-FX in my PCIe equipped desktop (upgrading from the HDSPe MADI) wink  I couldn't imagine the logistics of dealing with 64 tracks of MADI on a 13" Laptop, much less 3x that with the MADI-FX!  Otherwise, the USB UCX is likely ample for Laptop (and iMac systems based on compact Laptop components) applications...

cool

Spoken like a PC guy smile  Apogee has successfully made a lot of dough selling its Ensembles and Duets without ever even making them PC compatible.

Apple giveth and Apple taketh away. My guess is that PCI-e on the way out for Macs, period. I got rid of an older, slower Mac Pro Quad 2.66 n replaced it with a Mac Mini Quad Core i7 that runs circles around it for $1125, hooked up to a $1600 i7 PC slave I built and the tandem is very powerful and left me poised for the future.. My days of buying $5000 Mac Pro towers are over and a lot of Mac guys feel this way. And Mac or PC, make no mistake about it, laptops and smaller machines are where things are headed and the days of big towers with PCI-e slots are numbered IMHO.

UA, Avid, and others are already developing TB audio interfaces so clearly that "niche Market" is perceived as viable by them. And catering to "niche markets" can be quite profitable, asApogee has demonstrated.

I believe, there is a market for an RME TB interface and until RME can say to me, "Jay, those huge Logic Pro projects you are running will run at an equally low buffer, with equally high track counts and equal stability" and not a vague"close to" I think it would be wise.

But of course that is RME's call to make. I sure do love my HDSpe-AIO. It is a great interface.

OK, I am done.

Composer, Logic Certified Trainer Level 2
author of "Going Pro With Logic Pro 9"
www.jayasher.com

23

Re: Thunderbolt USB

I woud like to point out again that using an existing (or an upcoming) Thunderbolt to USB adapter the UCX and others are turned into a valid alternative to a unit that has TB directly inside. They are small, cost less than 200 bucks, and do not need power. Adding TB would significantly raise the costs for everyone, even the high percentage of users who will never use it. This might change next year, when Intel has cheaper chips and is willing to give them out to more than a handful of companies...

Regards
Matthias Carstens
RME

Re: Thunderbolt USB

Ashermusic wrote:
Randyman... wrote:

The Sonnet Express Pro chassis is like $799 - so it's only $300 more than what you're suggesting RME offer for $500 (and with less overall flexibility compared to an "open" PCIe chassis, and with a very small market of potential buyers).  And as you say, the Sonnet Chassis also opens up these limited "PCIe-less" devices to the PCIe standard as a bonus (not just an RME interface over TB, but any PCIe card you might need access to).  I'd say that benefit alone is worth the extra $300 for the Sonnet PCIe chassis.

iMac, MacMini, and Macbooks are all based around "compromised" laptop chipsets (they favor low-power draw over raw performance).  If you really need the INSANE bandwidth avaiable over PCIe (or TB) for Audio duties, you are likely better off with a desktop and its "full sized" chipsets and faster CPU's IMNSHO.  Otherwise, the USB UCX is (again) likely within 1ms of a TB system (round trip) for any of the "PCIe-less" Apple devices you mention.  Would you pay $500 (likely much more) for 1ms improvement in latency (USB UCX vs an RME TB Interface)?  That's a pretty small niche market IMO - and I can understand RME's reluctance on making a dedicated TB interface for the time being.

Maybe RME will re-consider this in a few more years once TB has gone through a few generation upgrades and has a larger install-base (as in: widespread PC adoption) and lower upfront cost (chipset and cable costs can come down quite a bit IMO).  In the meanwhile, I'll be more than happy to get my hands on the HDSPe-MADI-FX in my PCIe equipped desktop (upgrading from the HDSPe MADI) wink  I couldn't imagine the logistics of dealing with 64 tracks of MADI on a 13" Laptop, much less 3x that with the MADI-FX!  Otherwise, the USB UCX is likely ample for Laptop (and iMac systems based on compact Laptop components) applications...

cool

Spoken like a PC guy smile  Apogee has successfully made a lot of dough selling its Ensembles and Duets without ever even making them PC compatible.

Apple giveth and Apple taketh away. My guess is that PCI-e on the way out for Macs, period. I got rid of an older, slower Mac Pro Quad 2.66 n replaced it with a Mac Mini Quad Core i7 that runs circles around it for $1125, hooked up to a $1600 i7 PC slave I built and the tandem is very powerful and left me poised for the future.. My days of buying $5000 Mac Pro towers are over and a lot of Mac guys feel this way. And Mac or PC, make no mistake about it, laptops and smaller machines are where things are headed and the days of big towers with PCI-e slots are numbered IMHO.

UA, Avid, and others are already developing TB audio interfaces so clearly that "niche Market" is perceived as viable by them. And catering to "niche markets" can be quite profitable, asApogee has demonstrated.

I believe, there is a market for an RME TB interface and until RME can say to me, "Jay, those huge Logic Pro projects you are running will run at an equally low buffer, with equally high track counts and equal stability" and not a vague"close to" I think it would be wise.

But of course that is RME's call to make. I sure do love my HDSpe-AIO. It is a great interface.

OK, I am done.

Yes, that's the mac way you chose. Apple slims down hardware options in notebooks (e.g. no expresscard slot in MBP 15), and there is no alternative around. Solution: Become a PC user...

Re: Thunderbolt USB

What about USB3? Will it improve the capabilities of future audio interfaces significantly? At least it is available everywhere.

26 (edited by Randyman... 2012-09-11 07:40:29)

Re: Thunderbolt USB

MC - Are you suggesting that using a UCX (USB) via a Thunderbolt-to-USB dongle would somehow surpass using the Intel Chipset's native USB Ports on Intel based computers?  I assumed the native Intel USB Ports were generally the best choice for the UCX, and I also assumed all Intel based computers have at least a few Intel USB Ports available.

RE Ulrich: I don't believe USB3 would offer any appreciable improvement over USB2 with an audio interface - but I might be wrong (maybe a few percent CPU savings and 0.25ms RTL improvement with a highly optimized driver?)...

Regardless - the fact that RME's USB 2.0 performance is so darn impressive further negates the added cost of incorporating TB on such a device IMO.  $500 for ~1ms less latency and a few more plug-ins (due to TB's more efficient use of CPU) just doesn't make sense when RME's USB devices are already doing so well.  Then - Chances are if you really need more than the UCX's 18 Inputs/Outputs, then the 18x24 I/O of the Apollo (or 32x32 of the Apogee for that matter) isn't going to cut it, either.  PCIe (or PCIe-to-TB in "PCIe-less" systems) will still be required in the end IMNSHO...

cool

MADIface-XT+ARC / 3x HDSP MADI / ADI648
2x SSL Alphalink MADI AX
2x Multiface / 2x Digiface /2x ADI8

27 (edited by Ashermusic 2012-09-12 15:39:43)

Re: Thunderbolt USB

MC wrote:

I woud like to point out again that using an existing (or an upcoming) Thunderbolt to USB adapter the UCX and others are turned into a valid alternative to a unit that has TB directly inside. They are small, cost less than 200 bucks, and do not need power. Adding TB would significantly raise the costs for everyone, even the high percentage of users who will never use it. This might change next year, when Intel has cheaper chips and is willing to give them out to more than a handful of companies...

Ok, this has been interesting and has caused me to rethink my assumptions. I am certainly not going to argue digital audio with Matthias.

Composer, Logic Certified Trainer Level 2
author of "Going Pro With Logic Pro 9"
www.jayasher.com

Re: Thunderbolt USB

Bump for input from the man himself :-)


MC wrote:

I woud like to point out again that using an existing (or an upcoming) Thunderbolt to USB adapter the UCX and others are turned into a valid alternative to a unit that has TB directly inside. They are small, cost less than 200 bucks, and do not need power.

Randyman... wrote:

MC - Are you suggesting that using a UCX (USB) via a Thunderbolt-to-USB dongle would somehow surpass using the Intel Chipset's native USB Ports on Intel based computers?  I assumed the native Intel USB Ports were generally the best choice for the UCX, and I also assumed all Intel based computers have at least a few Intel USB Ports available.:cool:

I assume the "overhead" USB CPU utilization would still be handled by the system's CPU (and not the TB-to-USB dongle - unless tjhe TB dongle has some DSP or processor/translator of its own)?  Wouldn't this UCX USB > TB Dongle > TB Port configuration still use the RME USB driver (how else could ut work)?  Basically the same as using a PCIe USB card for a UCX?

Thanks for clearing up any misunderstanings I might have cool

MADIface-XT+ARC / 3x HDSP MADI / ADI648
2x SSL Alphalink MADI AX
2x Multiface / 2x Digiface /2x ADI8

29

Re: Thunderbolt USB

Hi Randy,

sorry if I have been unclear. Of course the UCX (etc) would operate like at a normal USB port and use its normal drivers with the TB adapter. Still it could work a bit better (read stable/reliable) as it can use a single USB port connected directly to a PCI Express bus - alone on its own and undisturbed, so to say.

Regards
Matthias Carstens
RME

Re: Thunderbolt USB

Makes sense to me!  I might even say leave the native "built in" Intel USB port(s) for the UCX, and then use the Thunderbolt dongles for your external HD's/Displays and other peripherals, as who knows what USB chips these cheap TB-to-USB adaptors are using ;-)

Thanks for confirming I wasn't goiong crazy! cool

MADIface-XT+ARC / 3x HDSP MADI / ADI648
2x SSL Alphalink MADI AX
2x Multiface / 2x Digiface /2x ADI8

Re: Thunderbolt USB

When RME say their current devices are USB 3.0 compatible that doesn't mean they will offer USB 3.0 performance does it? Am i right in thinking USB 3.0 performance would require a hardware upgrade rather than something that can be implemented via firmware etc?

I wonder how long FW interfaces are going to stay compatible with thunderbolt. FW seems to be well on the way out.

Re: Thunderbolt USB

Hello guys,
You seam really aware about TB, so would you help me please ? I need to change my old 1.1 macpro for a faster one. I plan to buy the new 2013. I'm also changing my Apogee FW ensemble because of latency in big logic pro sessions. I use guitar amp simulations, and I often have to move back audio files even with plugin compensation enabled. I had no such problems with my old RME PCI card + ADI 8. With thunderbolt I'm wondering what audio interface I should buy to get the best low latency ? if get the Raydat + ADI converter + sonnet PCIe chassis, will I'll same low latency as Symphony i/o + thuderbolt bridge ?

thanks !