Apogee Duet 2 drivers are bad bad bad!

[TNM]

Originally Posted by maxwelldub

Most definitely. Do u want my lowest usable latency under load? I use a lot of vstis and cpu heavy fx...

And i need to wait until someone buys my uln-2, which is in the classifieds btw

oh YES please the lowest usable under load would be GREAT.....

Also since you have a ULN-2.. that is one i would also LOVE to know figures fo.. how come you selling?

[bafonso]

Originally Posted by maxwelldub

Most definitely. Do u want my lowest usable latency under load? I use a lot of vstis and cpu heavy fx...

And i need to wait until someone buys my uln-2, which is in the classifieds btw

Wait, you're going to sell your uln-2 so you can buy the rme babyface?

[rifftrax2]

Originally Posted by Clawson

Also, make sure the USB cable is a USB 2.0 cable, not all USB cables are 2.0 compatible. The cable that came with Duet is of course USB 2.0, if you switched cables, make sure it's the correct cable.

Any usb 1.X/2.0 cable will have the exact same 24/28awg twisted pair conductors enclosed in an additional shield and exact same pinout. Hate to call you on your B.S. but that's purely a marketing myth. You'll get the exact same waveform on an oscilloscope if looking in the mhz range for any usb cable unless it's usb 3.0 which uses multiple sets of twisted pair conductors per termination.

[Clawson]

Originally Posted by rifftrax2

Any usb 1.X/2.0 cable will have the exact same 24/28awg twisted pair conductors enclosed in an additional shield and exact same pinout. Hate to call you on your B.S. but that's purely a marketing myth. You'll get the exact same waveform on an oscilloscope if looking in the mhz range for any usb cable unless it's usb 3.0 which uses multiple sets of twisted pair conductors per termination.

USB 1.x cables are not all the same quality and they were designed for data rates of up to 12Mbs. USB2.0 runs at 480Mbs. At rates that high we are not only talking about awg and twisted pairs, we are also talking about characteristic impedance. You will actually need a pretty advanced scope to see the effects between USB 1.x and 2.0 cables at speeds like that.
What it really comes down to is this... we have seen some failures with USB 1.x cables, so therefore we recommend useing cable that has been USB 2.0 approved.
Our goal is to give you recommendations for setups we know that work.
My intension of the post was to help trouble shoot for those having problems... not sell USB 2.0 cables.

Chris Lawson
Apogee Electronics

[Clawson]

Available immediately, Apogee has released a new Duet 2 Software Installer.

Who Should Update

All Duet 2 users should update to OS X 10.7.4 Lion to resolve Duet 2 recognition issues.
Also compatible with 10.6.8
Duet 2 users should update to this firmware and software version for the latest Apogee Maestro enhancements.
For those who installed the Interim Duet 2 Software Installer, this release re-establishes operation at 176.4-192kHz sample rates.

New Feature

Speaker outputs may be switched between a +4 dBu and -10dBV reference standard. The output meters now display the level of hardware outputs.

Please read the Release Notes included in the download for update instructions.

Apogee Electronics > Downloads

[Jason Hollar]

Hi all - first post here - I've been following this thread with great interest. This info from Apple and Apogee is good news. I hope everyone who's had duet 2 troubles can get the upgrades. cheers, jason.

[SoundEng1]

Installed updates all working great here

[lllubi]

hot plugging works on both ends immediately
looks fine till now

[maxwelldub]

Originally Posted by bafonso

Wait, you're going to sell your uln-2 so you can buy the rme babyface?

Yeah, its kinda big and I dont need 4 outs

[webstersp]

Have they realeased better drivers for the duet 2?
I had to return mine due to drop outs and my imac 27 had problems
recognizing it. had to unplug it, replug etc!!

And what about RTL of the duet2 vs babyface?

[Izzi]

Originally Posted by Clawson

Available immediately, Apogee has released a new Duet 2 Software Installer.

Who Should Update

All Duet 2 users should update to OS X 10.7.4 Lion to resolve Duet 2 recognition issues.
Also compatible with 10.6.8
Duet 2 users should update to this firmware and software version for the latest Apogee Maestro enhancements.
For those who installed the Interim Duet 2 Software Installer, this release re-establishes operation at 176.4-192kHz sample rates.

New Feature

Speaker outputs may be switched between a +4 dBu and -10dBV reference standard. The output meters now display the level of hardware outputs.

Please read the Release Notes included in the download for update instructions.

Apogee Electronics > Downloads

Just updated... What's up with the I/O name change?

Clawson, can you chime in here?

[NBarnes21]

How about a Duet 1 update???? Horribly spotty use with 10.7.4 here!

[ttown23]

After all of the duet 2 problems I've seen on this forum, you couldn't give me one for free at this point!

[edwinhurwitz]

Originally Posted by ttown23

After all of the duet 2 problems I've seen on this forum, you couldn't give me one for free at this point!

I had one and loved it. I just needed digital i/o, so I sold it, but it was a joy to use. For free, I'd take 5!

[ttown23]

Originally Posted by edwinhurwitz

I had one and loved it. I just needed digital i/o, so I sold it, but it was a joy to use. For free, I'd take 5!

Well, that's good to hear.

As a professional producer though, the headache with these units and the negative impact that would have on my workflow would obviate any use I would have for one of these, even if it were free

[scruffydog]

Just dropped one today!...that is purchased one..to replace my 'one'...that is my a
Apogee one.....words aren't enough to describe what a fine device this is for the money....
all good in logic and Ableton with mac os 10.6.8 and the latest USB driver update...
so far so good...brilliant...a HUGE step up!.....'OBVIATE????' THAT

[ttown23]

Originally Posted by scruffydog

Just dropped one today!...that is purchased one..to replace my 'one'...that is my a
Apogee one.....words aren't enough to describe what a fine device this is for the money....
all good in logic and Ableton with mac os 10.6.8 and the latest USB driver update...
so far so good...brilliant...a HUGE step up!.....'OBVIATE????' THAT

Glad to hear it... maybe these new drivers have fixed things?

Yes - OBVIATE - as in, make unnecessary. Check Webster's if you need more info on the meaning of the word.

[scruffydog]

well...you explained well enough which rather obviates websters i guess....
the duet 2.....i think they got it good this time....what a relief!..phew...

[webstersp]

Would be a great thing if all the problems of the duet 2 are gone!!
I would buy another one for sure!
Can't beat the sound quality of the duet 2 in this price range or sure, and portability is a big plus!
But I'll wait a bit and see what other duet 2 users have to say!!
Because as good as it is, if it slows down my workflow I don't care how good it can sound!!

[orchestranova]

So...after all the b*tching on here...has the latest driver for the Duet 2 fixed things???

[lllubi]

Originally Posted by orchestranova

So...after all the b*tching on here...has the latest driver for the Duet 2 fixed things???

after 10 months of b*itching
yes

[Ray Sigmond]

Running smooth here

[SeeDub]

Originally Posted by xqtion

I have 2 problems with my Duet 2 and wrote Apogee support.
1) Not being recognized occasionally on startup that everyone else is seeing
2) Some of my Reason songs are unplayable due to statics and noise at 64 samples and these only have 5~7 virtual instruments. And when I switch the buffer to larger settings, it gets worse. Workaround is to set it at 128 or larger and quit and relaunch Reason. And this occurs only on some of my songs so there is no apparent reason. I can play some other songs with more instruments without any problems at 64 samples so this is not a CPU overload issue.

Apogee's response regarding issue #2 was that it's normal for USB2 interfaces to have problems at buffer size 32 ~ 64 samples.

And my response was "I can run the same problematic songs without any problems on Mbox 3 which is also an USB2 device"

And their response was "I think it is not unusual to have to set your software's buffer size to a different setting for different audio interfaces."

I would love to keep using my Duet 2 but Apogee support's attitude is killing me. Instead of "It must be inconvenient that these issues are occurring let me help you and look into it". They're like "Uh~ there is no problem with our drivers that's the expected behavior". And after I made some arguments, they finally said "ok send me your Apple profile".

Come on Apogee, you guys are not giving your customers the trust and faith that you guys should be earning...

One big thing -- Sample Rates -- I have a Symphony and run it via USB -- if I use 96K - I have to set a large buffer as the math dictates. What I mean is this, at high sample rates a low buffer yields a lowest possible latency. At 44.1 you can go down to a real low buffer for USB. If you go up to 96K -you have to increase the buffer and still get the same total latency.

Are you using the same sample rate with MBox3 as with the Duet 2?

[spiral]

I thought i was fixed but i'm still getting the "won't boot" issue when hot-plugging. Only a reboot solves it after the Mac has slept a few times.

[TNM]

have the latest drivers helped with low latency performance at all?

[SeeDub]

Originally Posted by TNM

have the latest drivers helped with low latency performance at all?

Ok - I have been following this thread and am simply surprised at your problems with Duet mostly concerning the issue of latency. I was an extreme early adopter of native recording and dumped my Pro Tools system probably a couple years to early (RISK days, PPC, OS8-9) but since OSX and all intel based macs I found that latency in a native system became very manageable even when you where using software monitoring.

The exceptions for me are rare and usually very large DSP dense sessions using CPU hungry third party software instruments using large samples and/or I had become fixed on some of Logic's very well done mastering plug-ins (Linear Phase EQ - Adaptive Limiter etc). I realized there where limits that needed attention to detail to manage latency while monitoring through software. The solution in this situations was using the Low Latency monitoring functions of nearly every audio interface in the industry. In those early days - the RME Nubus solution saved me as their Low Latency monitoring was excellent.

Here is a quick detail list of what will affect latency

High Sample rates - you need to know that 96K at 32 samples in Logic is ultra low latency that takes some 'juice' to produce. This will give you a 3.6 ms round trip latency - that's really low for USB. At 44.1 at 32 samples it would be 4.3 ms. Still really low. With USB anything around 5 is excellent. If you can hear 5ms - then you need to use the interfaces low latency mode.

Mastering Plug-ins in Logic - when tracking mute or don't use plugins on your output and master faders.

Third Party Plugins and Instruments - some of my favorite companies (such as NI) are notorious for really poor buffer performance. Because playing software instruments is the one aspect that a low latency setting will not assist, this is my biggest pet peeve. In some RARE situations I have had to freeze tracks to get back some headroom.

Old CPU's - using old OSX. It has to be 10.6.3 or later if you are using USB. It has to be an Intel based Mac. Simple. This is all common sense in my opinion

Low Latency Setting. This is simply routing the audio input back to the output of an audio interface before the software. Most interfaces have it. The only drag with this setting is you should turn off software monitoring and then the signal is dry. There is a simple fix for this in Logic. Simply leave software monitoring on and when you set up a fader for tracking don't insert effects but rather use sends to aux faders with effects inserted there. Turn off the tracking faders output (temporarily) and then track. You will heard the audio input via the device (sub 1ms latency) but the software tracking fader will still have the sends active. The returns from the Aux's going to your outputs. All latency to the FX's via the sends will now be pre-delay. There are other ways to set this up that are a tiny bit slicker but I think you get the idea. I use this rarely. My favorite thing about Pro Tools HD Native etc.. it that this style of low latency tracking is seamless in the system. Avid go it right with their PCI systems.

As you can see, I really dislike the latency myths and mistakes out there, there are limits, there are reasons and there are solutions. You may need a new CPU or you may have a broken interface and you may need to examine how you are tracking and with what tools. All I know is the Duet is a pretty excellent box and with a 2X2 box, latency should be an afterthought and a non issue.

[mpmusicny]

SeeDub -

Do you actually own a Duet or a Duet 2? If so, I'd love to know how you get it to work at 32 or 64. For a lot of us of us, this great sounding box is unusable at these settings. It's not a "myth".

Originally Posted by SeeDub

Ok - I have been following this thread and am simply surprised at your problems with Duet mostly concerning the issue of latency. I was an extreme early adopter of native recording and dumped my Pro Tools system probably a couple years to early (RISK days, PPC, OS8-9) but since OSX and all intel based macs I found that latency in a native system became very manageable even when you where using software monitoring.

The exceptions for me are rare and usually very large DSP dense sessions using CPU hungry third party software instruments using large samples and/or I had become fixed on some of Logic's very well done mastering plug-ins (Linear Phase EQ - Adaptive Limiter etc). I realized there where limits that needed attention to detail to manage latency while monitoring through software. The solution in this situations was using the Low Latency monitoring functions of nearly every audio interface in the industry. In those early days - the RME Nubus solution saved me as their Low Latency monitoring was excellent.

Here is a quick detail list of what will affect latency

High Sample rates - you need to know that 96K at 32 samples in Logic is ultra low latency that takes some 'juice' to produce. This will give you a 3.6 ms round trip latency - that's really low for USB. At 44.1 at 32 samples it would be 4.3 ms. Still really low. With USB anything around 5 is excellent. If you can hear 5ms - then you need to use the interfaces low latency mode.

Mastering Plug-ins in Logic - when tracking mute or don't use plugins on your output and master faders.

Third Party Plugins and Instruments - some of my favorite companies (such as NI) are notorious for really poor buffer performance. Because playing software instruments is the one aspect that a low latency setting will not assist, this is my biggest pet peeve. In some RARE situations I have had to freeze tracks to get back some headroom.

Old CPU's - using old OSX. It has to be 10.6.3 or later if you are using USB. It has to be an Intel based Mac. Simple. This is all common sense in my opinion

Low Latency Setting. This is simply routing the audio input back to the output of an audio interface before the software. Most interfaces have it. The only drag with this setting is you should turn off software monitoring and then the signal is dry. There is a simple fix for this in Logic. Simply leave software monitoring on and when you set up a fader for tracking don't insert effects but rather use sends to aux faders with effects inserted there. Turn off the tracking faders output (temporarily) and then track. You will heard the audio input via the device (sub 1ms latency) but the software tracking fader will still have the sends active. The returns from the Aux's going to your outputs. All latency to the FX's via the sends will now be pre-delay. There are other ways to set this up that are a tiny bit slicker but I think you get the idea. I use this rarely. My favorite thing about Pro Tools HD Native etc.. it that this style of low latency tracking is seamless in the system. Avid go it right with their PCI systems.

As you can see, I really dislike the latency myths and mistakes out there, there are limits, there are reasons and there are solutions. You may need a new CPU or you may have a broken interface and you may need to examine how you are tracking and with what tools. All I know is the Duet is a pretty excellent box and with a 2X2 box, latency should be an afterthought and a non issue.

[mpmusicny]

And with regard to the limits of the system, there is a healthy debate about the various sonic attributes of the duet 1/2 against the Babyface, but one fact that does not seem to be contentious is that the Babyface is way better at functioning at low latency settings. Which leads to the conclusion that Apogee's code is at fault. Again, I love the sound of the thing, and apart from the low latency non-show it works as advertised in snow Leopard, but i have been suspicious of the limits of their software prowess ever since they failed to make the original Duet Aggregate-able . .

[SeeDub]

Originally Posted by mpmusicny

And with regard to the limits of the system, there is a healthy debate about the various sonic attributes of the duet 1/2 against the Babyface, but one fact that does not seem to be contentious is that the Babyface is way better at functioning at low latency settings. Which leads to the conclusion that Apogee's code is at fault. Again, I love the sound of the thing, and apart from the low latency non-show it works as advertised in snow Leopard, but i have been suspicious of the limits of their software prowess ever since they failed to make the original Duet Aggregate-able . .

I have a Duet 2 - and those settings came from my Mac Book Pro and Logic 9.1.7 - The driver is not Apogee's - the USB driver is Apple's class compliant USB audio driver (updated in 10.6.3 - Snow Leopard). The only thing that is installed from Apogee is the control and Maestro software. Aggregate driver is a no fly zone in my opinion. Use at your own risk. The Babyface could in fact NOT be using the class compliant driver and they could be doing some cool stuff with a proprietary driver. Remember I am an RME fan and they where my go to audio interface company before Apogee. I am all for choosing a solution based on driver performance. If you have at least 10.6.3, a newer intel based Mac, at least 8gb ram (4gb gets testy), follow some of the common sense stuff I mentioned - then you should he able to use the Duet at 64 sample buffer at 44.1 in Logic without much fuss -- at 128 - it's still 8.7 ms. I can play software instruments fine at 128. 256 even. For monitoring vocals - the Low Latency mixer is your friend.

I only offered an opinion and solution and a possible baseline measurement to compare systems and what I have experienced as reasonable. Again - I am all for choosing any interface that has superior driver performance. I would be ok with a hit in fidelity to get better driver performance if all of the things I mentioned did not work. Just trying to help - and frankly if super performance and latency are the object du jour - you should be using a PCI solution as that is still the best data path for audio.

The RME stuff is great - I just have no issue with USB and the Duet 2 -

[spiral]

I'm thinking of getting a Scarlett 2i2 or MOTU Hybrid mk3. Have any Duet 2 owners gone that route in light of the driver weirdness? If so, was the quality, and performance (latency, stability, etc.) acceptable to you? I only need a few channels of conversion for line level sources.