Some new driver info from MOTU

[monkey man]

So, Steve, you don't have to save up for a nMP and all the attendant peripheral nonsense. Your 2009 unit should be good to go with that card, but I'd check this by asking the question here first to be certain as I've only confirmed this for my 2012 model

Nicky, are you saying you are running AVB on your tower, or planning to? If this card works, my interest in AVB will escalate.

Yep, Billy, planning on it. I've been aware of the AVB option for about 15 months I think (previously was going to use USB), and as I said, learning of it meant I could cease stressing about which synths to seek in order to meet the USB channel-count restriction. The latency put me off for a further 6-9 months, but 6 months ago I decided I don't care about the extra 5ms 'cause found the prospect of keeping all my synths a far weightier attraction. Here's the thread I started 15 months ago where the AVB-connection option was first discussed here. Interesting reading, but probably a little(!) confusing overall:

http://www.motunation.com/forum/viewtop ... =2&t=59063

Things have crystallised quite nicely in this thread in comparison to that one.

This whole "old-Mac" thing sure did cause me a lot of stress since July 2014 when the AVB line first appeared. I posted quite a lot here on the subject of our being "forced" to update our Macs in order to be able to run decent channel counts; I wasn't a happy Kemper. It would've taken many years to save the dough for the peripherals and helmet and I couldn't understand why MOTU would "tease" us this way.

Fortunately, the e-card option provides a way out, albeit a little imperfect:
Latency is 8-9ms, and system-audio volume, as is the case with USB, cannot be controlled via the usual keyboard buttons. I'm still not clear on whether or not it can pass through an AVB interface, in which case one would have to dedicate external-mixer channels for the Mac's line outs. If indeed it is passed, the external-mixer channels would still be required but would receive a higher-quality signal from specified AVB-unit outputs.

Billy, I'll expand greatly on this stuff in my response to HC, some of which is intended for you as well, which follows. Take care bud.

Cheers mate.

Thanks for the info on the availability of an AVB-compliant ethernet card for the cMP (cheesgrater/classic Mac Pro) MM. I had been wondering about that, and had begun researching, only to see your on-point post here at MOTUNation. Great news for continuing the useful life of the cMP!

You know, I'm glad to have been able to cut your research short, Sir Markus.

My Googling yielded nothing but pain. The only candidates I found were the Privitec CMx64io-AVB Network Card, which is a CM-1/CM-2-format one, and the Privitec Streamware NIC-1, which is, er... $800 US for an itty-bitty PCIe card, the smallest I think I've ever seen, sporting the fewest chips I've encountered too:

http://echoavb.com/products/streamware-nic-1#tech-specs

Sir Markus, this is also for Billy and others interested too:

The Sonnet one I linked to in my previous post retails for around $100 US and is on special at Amazon right now for $39 US with free shipping (3 left):

http://www.amazon.com/Presto-Gigabit-Et ... B0017K6BR8

This one's $49 with free shipping (35 left):

http://www.amazon.com/Sonnet-Technologi ... B00IA94IRY

Judging by the user reviews on those pages it's compatible with early-2008 MPs onwards.

WARNING: I've a suspicion the original chipset wasn't ideal for / didn't work with the MOTU gear, in which case the Revision-B model would be in order. This card apparently is compatible with "early 2008-2012" units with PCIe slots according to this vendor. The price is in $AU:

http://www.epowermac.com.au/Shop/pc/Son ... -p3241.htm

I see no differentiation between the two models on the Sonnet site; the previous model has been replaced on the same page. The specs, however, now say "GE1000LAB-E Presto Gigabit Pro PCIe (Revison B model)". That part number, I'm guessing, is the key to ensuring you have the right card:

http://www.sonnettech.com/product/prestogigpciepro.html

I've checked that the Amazon stock, at least, that on the links I provided, is indeed the newer model, but felt this heads-up was in order in case you'd like to confirm somehow (maybe upon checkout?) that you've the correct one. The main pages quote the correct part number, but as I suggested, it might be worth double-checking if possible.

How is MOTU doing on bringing latency down on the interface to computer direct ethernet connection? Last I heard, it was higher than for USB2.0, which is higher than for Thunderbolt.

Correct, HC.

1-2ms, 4ms and 8-9ms respectively, from memory.

Mr. Miller told me late last year that there's nothing more MOTU can do about the e-net figure; it's up to Apple now and progress will depend upon the development of its implementation into the future.

Given MOTU's promised halving of latency at any given buffer setting (in DP, revealed at NAMM) relative to what we're used to, I'm counting on this as being able to compensate sufficiently. I'd have to buy DP9 I think. Also, considering the lack of DHP, many users, I expect, would only be affected by the latency figure when playing VIs. In this scenario, system-throughput latency (including DP but not the e-card) would theoretically be halved again as it's only a half-trip for the audio, plus of course the MIDI-triggering latency.

Phew! I think that just about covers it. Hopefully my lack of sleep translates to softer pillows for you guys!

Take care now...

[bayswater]

1-2ms, 4ms and 8-9ms respectively, from memory.

Can you use more than one connection at a time? E.g. could you set it up so channels you track on are via USB with 4ms latency, while others are at 8-9ms?

Not that 8-9 is that bad. Same as standing 9 feet from the stage monitors or your guitar amp. Most of us that still have hearing have managed that.

[monkey man]

Whilst I believe it is possible to use two connections (USB for the single unit and, say, AVB or TB for the remaining "pool"), I've a strong feeling that one cannot then run them all on the same network. IOW, AVB routing through a switch box, or even an e-net connection between two or more units, won't allow shuttling of audio between the interfaces. This would obviously mean not being able to bring all one's input sources into, say, a stereo pair or two a-la CueMix for monitoring.

If you don't require this and are happy to marry all the non-USB-connected interfaces into a network, connect them via an AVB card (from MOTU's switch box) and monitor them the CueMix way, as you might've done otherwise, and run a discreet / "standalone" interface over USB "on the side" for the lower-latency performance you asked about. I can see no reason why that wouldn't work. The downside is that this unit would have to be brought into an analog monitoring mixer alongside your AVB-network stems from the other units, which would effectively be behaving as a single entity. Unless, of course, you're happy to monitor through DP, taking advantage of the 5ms-reduced latency.

So, two empty-titties, two cables and possibly at least one extra stereo pair to be mixed in the analogue domain, or two empty-titties, two cables and live monitoring through DP along with the attendant latency induced by your buffer settings.

You could of course mix the discreet unit's output into the AVB network by physically patching it into spare inputs (audio), but seeing as you're asking about monitoring through the 'puter, there'd be no point.

I'm sufficiently confused and have waxed lyrically enough on a subject that does my head in big-time, so I'll cease spectacular-titty-lactating right here... now.

Cheers Stoivo.

[billf]

Yep, Billy, planning on it. I've been aware of the AVB option for about 9 months I think (previously was going to use USB), and as I said, learning of it meant I could cease stressing about which synths to seek in order to meet the USB channel-count restriction.

This whole "old-Mac" thing sure has caused me a lot of stress since July 2014 when the AVB line first appeared. I posted quite a lot here on the subject of our being "forced" to update our Macs in order to be able to run decent channel counts; I wasn't a happy Kemper. It would've taken many years to save the dough for the peripherals and helmet and I couldn't understand why MOTU would "tease" us this way.

Fortunately, the e-card option provides a way out, albeit a little imperfect:
Latency is 8-9ms, and system-audio volume, as is the case with USB, cannot be controlled via the usual keyboard buttons. I'm still not clear on whether or not it can pass through an AVB interface, in which case one would have to dedicate external-mixer channels for the Mac's line outs. If indeed it is passed, the external-mixer channels would still be required but would receive a higher-quality signal from specified AVB-unit outputs.

This is certainly welcome and promising news! I look forward to how well it works in practice once you have it going.

[Sean Kenny]

P

[monkey man]

Hmm... I'm not an expert on internet speak, Sean.

Does "P" mean "promising"?

[bayswater]

Does "P" mean "promising"?

You're asking? P can only mean one thing ...

[monkey man]

Ha ha ha! Good one, Stoivo.

I'll call the MonkeyLabs Industries KleanUp Krew™ (Any poo, any time, anywhere) now...