Saffire PRO 40

This device is NOT supported in 2.0. It was included in the 2.1 release that introduced DICE support.

A user (Jannis Pohlmann) has reported that a firewire host interface based on the VIA VT6315 chipset is not suitable for running this device at 88.2 kHz. The Pro40 worked fine with the same computer when a Texas Instruments XIO2213 based firewire card was used instead.

Damien has provided a tutorial which walks through his experiences with this device. It's currently available only in French: Audio details, Using ffado-mixer.

Support Status: 
Reported to work



this is probably a dumb question, but I just compiled ffado and jack from svn and all seems to work (recognizes my Saffire Pro 40). But where is the mixer app? After scons/scons install, there doesn't seem to be a ffadomixer executable anywhere and I cannot seem to be able to start any of the python files in the source directory. Also, I cannot find any mention of it in the compile output (like a warning that something is missing).


i'm wondering to myself the same question.
It was once (i dont remember waht revision number) the mixer worked.
Just I think ffado is under developing and some day I will run the mixer again without "errors".
It will be.

These are strange questions.

If you can't find any ffado-mixer after building from source, you are missing a dependency. Fix it the real way, not with some hacks indicated elsewhere.

Other than that ffado-mixer (mostly) works here for the pro24 and while I know the differences to the pro40, I don't have direct access to that device, so I can't test immediately. But if you encounter problems, please create tickets (after checking whether that issue wasn't reported before).

Yes ffado-mixer exist after a clean instalation from svn but with errors.

I've reported three issues. About ffado-mixer, playback at 96000 samplerate, dice_eap.cpp: No routes found

I don't know if i've created good and usefull tickets because my english is so bad.


Saffire pro 40, svn version from 1 week ago, ffado-mixer working out of the box: everything running quite well. I'm having problems to understand the "dice" mixer usage, I mean, how the matrix works, how to route from there (Or should I use jack straight??) and so forth.

Thanks in advance


The mixer is a matrix. So you have 16 mono-mixes from 18 inputs. Where the sound comes from and where it goes is defined in the router.

The dice-router can route any input to any output. But while one inputs signal can be routed to several outputs, one output can only get a signal from one input. Thats why in the router you see an item for each output and can select where the sound comes from. Note that the inputs of the mixer are outputs (destinations) for the router and the outputs of the mixer are inputs (sources) for the router.

That gui is not optimal yet, I am working on one that is (hopefully) easier to understand...

The problem is that the SConstruct file asks for a command "pyuic" (if conf.CheckForApp( "which pyuic" )...) which, on Ubuntu, is in the package python-kde3-dev. This suggests that it is related to QT3, whereas the package pyqt4-dev-tools only contains "pyuic4" which seems to be the same for QT4. I was under the impression that I should rather not let any QT3 tools and sources spoil anything, or should I install the kde3 package with "pyuic" anyway?


SConstruct searches for both. The 2.0 release still also contains the Qt3-based mixer (which doesn't get any love anymore and shouldn't be used).

When you install the pyqt4-dev-tools, scons should pick up the pyuic4 and built the Qt4-based mixer.

I was wondering to myself the same question.
But ffado 2.X is under development. It was some svn revision (i dont remember the number) than ffado-mixer worked ok.
I also get messages about error in dbus when I run ffado-mixer.
While we wait an operative mixer you can reinstall the ffado-mixer wich comes in ffado2.0rc2. From it at least you'll can change the sample rate and the clock source.
If you are working with some distros (Debian, ubuntu 9.10, Musix, ...), you can reinstall it (ffado-mixer-qt4 svn1569) from synaptic

Ok, I just realised that the instructions I followed here: seem to be from before a switch to QT4 and it builds when the replacement line is

env['PYUIC4'] = True


With the PRO 40 I can capture at this sample rate ( 96.000 )
But when I try to playback it I can hear just nothing.

It works OK with 44.1K and 48K.
Anyone knows what is the reason?


as ffado is telling you in a cryptic way, there is no default routing table for 96kHz. Or just an incomplete, where the inputs are routed to firewire and the firewire-output is routed nowhere. Or to a mixer that has all coefficients zero...

Take a look at ffado-mixer with the matrix-mixer and the routing table. After switching your device to 96kHz and restarting both ffado-mixer and ffado-dbus-server (the automatic switching isn't working yet).

When I change routing thru ffadomixer nothing happens or it happens someting wrong.
After I try to use ffadomixer the device vumeters strangely change. If I send a signal to input #2, the vumeter wich blinks is number 7 ( or any other).
And when I change somthing in ffadomixer (routing or matrix mixer) the data sometimes are stored sometimes not or routing changes itself after I restart ffado mixer.

After I try to use ffado mixer I need to reset the device with focusrite software under windows.

The mixer and router should work. At least it does for the Pro24 at 48kHz, I will test at 96kHz soon.

The settings made in the mixer are stored permanently in the device when ffado-dbus-server is shut down while the device is plugged in. There will soon be a gui-button to store this at any time from within ffado-mixer.

I am sorry that you have to run the windows software to restore the state, rest assured that here I don't (yet) face that problem...

I dont know why, but it doesnt work properly even at 48000.
The matrix and routing do what they want.
When I start ffadomixer and I change somthing in the routing the device vumeters change its inpput. As I described before.
And once I've changed some thing and I restart ffadomixer sometimes it remenbers my changes or sometimes not.

Every time I change routing, p.e. Analog:00 to 1394:00, ffado-dbus-server says:
11532126752: Warning (dice_eap.cpp)[1396] read: No routes found. Base 0x7, offset 0x4000
11532154895: Warning (dice_eap.cpp)[ 879] updateNameCache: What is this function about?

Maybe it can help you to fix it.

These are "only" warnings.

They are to remind me of things still to do. But these are not errors that stop your device from working...

Thanks and sorry arnonym I was mistaken inputs and outputs. Finally I was able to get routing correctly.
And the matrix also works OK.
The only strange thing is about the device vumeters (input 1 signals shows in VU 7, input 2 in VU 8 , ...) and sometimes VUs dont blink.

It is a great idea getting a button to save settings in hardware and maybe also would be great to save them in a file, because every time I start with windows the device resets its setup.

Thank you again

I've managed to get this working perfectly running at 48KHz, but at 96KHz I am having the problem described above - input finds its way into the computer fine, but output doesn't find it's way back out of the device. On the Saffire's front panel meters, the signal from input#1 is displayed on meter#7. Settings in ffado-mixer are apparently the same as when in 48KHz, and hardware monitoring works.

Can anybody point me in the right direction?
I'm setting this up for a mate who I convinced to run Linux (and the Saffire because of Focusrite's Linux support), so I'd really like it running in top form when I give it to him.

the problem is that ffado-mixer has to reload the routing when the sampling-rate is switched to 96kHz. Otherwise it will definitely show the same as with 48kHz, because in fact it still shows the setting of 48kHz. quit ffado-mixer and kill ffado-dbus-server and start ffado-mixer again. that should give you the real routing and mixer for 96kHz.

I haven't tested this yet...

But I know about the issue, its on the todo-list.

Alright, with r1794, the mixer seems to work. Thanks!

However, now I cannot start jackd:

mir@robotnik:~/Downloads/jack/jack$ sudo /usr/local/bin/jackd -R -dfirewire -r48000 -p1024 -n3 -i8 -o8
[sudo] password for mir:
jackd 0.118.2
Copyright 2001-2009 Paul Davis, Stephane Letz, Jack O'Quinn, Torben Hohn and others.
This is free software, and you are welcome to redistribute it
under certain conditions; see the file COPYING for details
no message buffer overruns
JACK compiled with System V SHM support.
loading driver ..
00753005682: (ffado.cpp)[ 92] ffado_streaming_init: libffado 2.999.0-1794 built Jan 18 2010 22:59:06
00765791063: Warning (dice_eap.cpp)[1396] read: No routes found. Base 0x7, offset 0x4000
00765841750: Warning (dice_eap.cpp)[ 879] updateNameCache: What is this function about?
firewire ERR: Could not start streaming threads: -1
DRIVER NT: could not start driver
cannot start driver
no message buffer overruns

Disconnecting and reconnecting the device (like suggested here) doesn't seem to help.

Sadly you are not showing the real source for the error.

Can you run jackd as user (otherwise you have to start _all_ jack apps as root), without the -i and -o flags, with -v6 and sent the output to a file with "&> ffado.log" and show that file? And please provide the output of "ffado-diag".

I'm getting this quite often as well, though 96khz seems to be working better. I'm having more trouble with the mixer/routing panel. Is there any documentation, I can't really manage to do correct routing on my own.

Im trying to setup my saffire pro 40, and am not having much luck. It seems that everything is installed, when i run "ffado-test ListDevices" my pro 40 shows up and seems to be communicating, but when i run "ffado-test Discover" I get the following

brucio@ubuntu:/etc$ ffado-test Discover
FFADO test and diagnostic utility
Part of the FFADO project --
Version: 1.999.43
(C) 2008, Daniel Wagner, Pieter Palmers
This program comes with ABSOLUTELY NO WARRANTY.

09979055244: Debug (devicemanager.cpp)[ 332] discover: Starting discovery...
09979078732: Debug (ieee1394service.cpp)[ 521] readNoLock: raw1394_read failed: node 0xFFC0, addr = 0x0000FFFFF0000480, length = 16
09979081216: Debug (ieee1394service.cpp)[ 521] readNoLock: raw1394_read failed: node 0xFFC0, addr = 0x0000FFFFF0000480, length = 16
09979083463: Debug (ieee1394service.cpp)[ 521] readNoLock: raw1394_read failed: node 0xFFC0, addr = 0x0000FFFFF0000480, length = 16
09979085713: Debug (ieee1394service.cpp)[ 521] readNoLock: raw1394_read failed: node 0xFFC0, addr = 0x0000FFFFF0000480, length = 16
09979088390: Debug (ieee1394service.cpp)[ 521] readNoLock: raw1394_read failed: node 0xFFC0, addr = 0x0000FFFFF0000480, length = 16
09979125751: Debug (ieee1394service.cpp)[ 521] readNoLock: raw1394_read failed: node 0xFFC0, addr = 0x0000FFFFF0000480, length = 16
09979128392: Debug (ieee1394service.cpp)[ 521] readNoLock: raw1394_read failed: node 0xFFC0, addr = 0x0000FFFFF0000480, length = 16
09979130764: Debug (ieee1394service.cpp)[ 521] readNoLock: raw1394_read failed: node 0xFFC0, addr = 0x0000FFFFF0000480, length = 16
09979133282: Debug (ieee1394service.cpp)[ 521] readNoLock: raw1394_read failed: node 0xFFC0, addr = 0x0000FFFFF0000480, length = 16
09979135643: Debug (ieee1394service.cpp)[ 521] readNoLock: raw1394_read failed: node 0xFFC0, addr = 0x0000FFFFF0000480, length = 16
09979196897: Error (ieee1394service.cpp)[ 702] doFcpTransaction: FCP transaction didn't succeed in 20 tries
09979196947: Warning (ieee1394service.cpp)[ 677] transactionBlock: FCP transaction failed
09979196958: Error (bebob_avdevice.cpp)[ 96] probe: Number of channels command failed
09979218176: Error (ieee1394service.cpp)[ 702] doFcpTransaction: FCP transaction didn't succeed in 20 tries
09979218198: Warning (ieee1394service.cpp)[ 677] transactionBlock: FCP transaction failed
09979239565: Error (ieee1394service.cpp)[ 702] doFcpTransaction: FCP transaction didn't succeed in 20 tries
09979239590: Warning (ieee1394service.cpp)[ 677] transactionBlock: FCP transaction failed
09979239616: Error (avc_avdevice.cpp)[ 88] probe: Subunit info command failed
09979239635: Debug (devicemanager.cpp)[ 639] discover: Discovery finished...
09979239656: Debug (devicemanager.cpp)[1184] showDeviceInfo: ===== Device Manager =====
09979239669: Debug (Element.cpp)[ 121] show: Element DeviceManager
09979239684: Debug (devicemanager.cpp)[1192] showDeviceInfo: --- IEEE1394 Service 0 ---
Iso handler info:
Dumping IsoHandlerManager Stream handler information...
State: 2
no message buffer overruns

I have read a ton of other posts and have followed all the steps, i am part of the video group which has access to raw1394, it is set to load on startup, but nothing seems to be working. Below is the error i receive when i try ti start jack. Any Help would be greatly appreciated.

17:29:10.729 /usr/bin/jackd -R -p128 -dfirewire -dhw:0 -r44100 -p128 -n2 -i4 -o4
17:29:10.735 JACK was started with PID=3600.
no message buffer overruns
jackd 0.116.1
Copyright 2001-2005 Paul Davis and others.
This is free software, and you are welcome to redistribute it
under certain conditions; see the file COPYING for details
JACK compiled with System V SHM support.
loading driver ..
10178429058: (ffado.cpp)[ 92] ffado_streaming_init: libffado 1.999.43 built Sep 17 2009 20:03:51
10178690502: [31mError (bebob_avdevice.cpp)[ 96] probe: Number of channels command failed
[0m10178732913: [31mError (avc_avdevice.cpp)[ 88] probe: Subunit info command failed
[0mfirewire ERR: Error creating FFADO streaming device
cannot load driver module firewire
no message buffer overruns
17:29:11.377 JACK was stopped successfully.
17:29:11.377 Post-shutdown script...
17:29:11.378 killall jackd
jackd: no process found
17:29:11.797 Post-shutdown script terminated with exit status=256.

Hi there. I'm very excited about the fact that Focusrite is interested in helping the comunity to create open source drivers for their hardware. I'm a linux user for quiet some time now and have worked with audio in linux since the times of ardour 1.X. But I have never put a finger in anything that have to do with FireWire, but I will make the move with a Saffire PRO 40. My question is, what about the ADAT and S/PDIF ports in the Saffire? have any one have success with ADAT or S/PIDIF? Also I didn't understand very clearly the mixer/router stuff. FFADO includes a mixer/router that actually controls the hardware?

thanks for the answer

pd. (sorry if I wrote to much... is the first comment I ever leave in my life!)

I'am using a TLAudio with spdif outputs towards PRO40 spdif inputs, and it works very well
At the begining I lost the sync, but after I link analog grounds it worked correctly ( I think it was TLAudio fault).
I've never tried ADAT port, but i think it would must work ok.


2.0 and 2.0.1 don't work for me, but svn appears to work from the computer side of things. I can't actually get any sound out of the interface though. I just tried the interface with a Mac and reset all the mixer settings to factory defaults, and I can get sound there.

I'm using an expresscard FW interface with a TI chipset:

03:00.0 PCI bridge: Texas Instruments XIO2000(A)/XIO2200(A) PCI Express-to-PCI Bridge (rev 03)
04:00.0 FireWire (IEEE 1394): Texas Instruments XIO2200(A) IEEE-1394a-2000 Controller (PHY/Link) (rev 01)

I get very occasional single LED blips on the VU meters, but actually tapping a mic plugged in does not affect them. The only errors/warnings I get are:

03275443083: Warning (dice_eap.cpp)[1398] read: No routes found. Base 0x7, offset 0x4000
03275472126: Warning (dice_eap.cpp)[ 881] updateNameCache: What is this function about?

I've been without a working audio setup for way too long now, and would really like to get back to LAD hacking. Please let me know anything (debugging, more info, hacking, money, anything) I can do to resolve this ASAP.


Its stated right above your comment that 2.0.X will not work for the saffire pro40.

Note that the LED-meters on the front are only to get a first impression. Did you actually try with running jack and connect meterbridge?

The warnings are to be ignored, the first means that there are no routes predefined for rates 88kHz and above and the second is just a reminder for me to look at that function...

Yes, as I said everything appears to work from the computer side of things. Jack (1) is working fine, meterbridge scope shows signal on the thing connected to the interface, etc. I just get no actual sound on any output, with a signal plugged in to any/all outputs on the interface.

Firmware is latest, mixer settings are stock.

There will be no sound output if you are using the new firewire stack. We are working on that...

I had no sound for about a day.
my setup.
ubuntu lucid, then upgraded to studio. uninstalled, reinstalled jack, then build 083 jack, then uninstalled and reinstalled from lucid repos 0711 jack.
Jack 0711 (whatever in ubuntu repositories) bunch of other updates.
ffado latest from the trunk.
gigabyte 8ig1000 2.66 1gb ram 2tb hdd,A texas instruments (bcm brand) pci dual firewire + 4usb card out of sunblade 1500, +annoying clicking pauses happening/ubuntu studio wireless.

Have an Alesis QS6, 4* shure pg drum mics, voc mics, guitar and fender prec bass through 15" fender amp (line out) in pro40, then out the monitor ports to 200w stero amp, then headphones.

Then it was my all my problem and learning to use jack/ardour.
Are the clicktrack 1/2s, the headphone outs?

I have sound out the montior ports and the outputs. But not headphones, not sure why.

Also the gains need to be turned way up to get sound out. Only realised that I had output, when I heard a strange deep buzzing from the speakers.

The mic on wooden desk was picking up the rumble and whirl of the PC fan & hdds.

Think I need quality headphones.

After months of not using my equipment, I'm giving this another try with revision 1905. I think this time I've compiled and installed everything correctly. But besides missing playback, I cannot record from a connected microphone, either. When I connect the microphone to the first input and meterbridge to capture_1 it shows a constant maximum (?) level, but not the signal I'm expecting.


Working via a Texas Instruments XIO2200(A) (in a Startech Expresscard) with Linux compiled with old Firewire stack.
New stack appears to "work", but with no actual sound I/O.

It also works in the new stack, trust me.

Only issue is that playback is muted in the new stack. The device receives the stream, but it isn't routed to playback. But recording works fine.

This problem exists with all Dice-based devices that use the Extended Application Protocol (EAP) and is being worked on.


Device have sync, is working pretty nice for recording, but i cant hear what i recorded ... wtf!
What can i do to put it to work?? Thanks for the info.

Im running:
Linux debian 2.6.32-5-amd64 #1 SMP Sat Oct 30 14:18:21 UTC 2010 x86_64 GNU/Linux

I couldn't get the output working with the newer firewire stack and was getting xruns.

Enabled the older firewire stack and the xruns went away and the output started working.


How do you change from the new to the old FW stack?

I'm using Debian Squeeze now; is there a way to select between one or another?



You need to have the appropriate stack enabled in the kernel you are running. I am not aware of any way of changing it on the fly (however I am certainly no expert on this).

If you can find someone's pre-prepared kernel with the features you desire (old firewire stack, realtime patch etc) as a .deb, that is the easiest way of getting up and running. For example there are a selection of kernels for AV-Linux here (these *might* work for you in vanilla Debian):

The hard-core way is to compile your own kernel, in which case you can enable/disable any feature you like. I like to use KernelCheck to help me with the process:
It will lead you through the process of downloading source code, selecting features, applying patches (realtime patch) and creating a .deb package. When selecting options you'll find a "firewire" section where you can enable the old stack (and disable the new one).
Read about applying the realtime patch here:

Once you are running a kernel with the old stack there is a couple of steps you'll want to take so that you don't need to run jack as root. My notes list these steps:
- Add "raw1394" to /etc/modules (to create the raw1394 device on startup)

- Create a udev rule for raw1394 (to set user/group permissions on startup)
Eg. In a new file /etc/udev/rules.d/51-firewire.rules place:
KERNEL=="raw1394", GROUP="audio", MODE="0664"

- Add user(s) to whichever group you've used in the previous step (audio, disk, video... whatever):
addgroup user audio

- Restart

Other notes:
- You must use a newer version of ffado than the 2.0 release (from SVN or otherwise).
- I have never got output/routing working at 96KHz (44.1 or 48KHz do work... a bit).

... Yes, it's certainly a bit of messing around - I can't wait until this device works properly on the new stack!

See the link posted above ( Here's how:

To select which firewire stack is loaded at boot time:

sudo gedit /etc/modprobe.d/blacklist-firewire.conf

By default in Maverick, the legacy stack is blacklisted and the Juju stack is enabled:

blacklist ohci1394
blacklist sbp2
blacklist dv1394
blacklist raw1394
blacklist video1394

#blacklist firewire-ohci
#blacklist firewire-sbp2

To enable the legacy stack, modify this file as follows:

#blacklist ohci1394
#blacklist sbp2
#blacklist dv1394
#blacklist raw1394
#blacklist video1394

blacklist firewire-ohci
blacklist firewire-sbp2

Then, rebuild the initramfs:

sudo update-initramfs -k all -u

Ubuntu 11.04 Natty uses the new stack, which as stated "appears" to work, but I get no output.

What I did was I installed Ubuntu 10.04 which uses the old stack, and added a newer version of ffado by adding a ppa that included it.
Haven't yet tested everything, but outputs are working fine.

After a few days of trying, I finally have a fully working Pro 40 in Ubuntu 10.10.
First off, as many have already said - this doesn't work yet with the new stack. So, if you want to save some frustration,
then load the old stack - and it should be a bit easier.

The first thing I did was load the subversion ffado drivers, which are already available in synaptic, so that was easy enough.
Same thing with jack. Once you get everything loaded and verify you're running the old stack, you should be good to go.

Here is the part that took numerous tries to get right. Make sure your Pro 40 is OFF when you boot up into the desktop.
Open a terminal and type "ffado-dbus-server &" but don't hit enter yet - power on your Pro 40 and after about 3 to 4 seconds hit enter to run ffado-dbus-server &
If ffado-mixer works you're good to go, if not - power off the Pro 40 and try the same steps again. I usually get it to work by the second try.
The trick it to get it recognized before it goes into standalone mode. I believe it has something to do with the Pro 40 only sending the device id right after it boots up?

Keep an eye on the terminal and you should see something like:

(ffado-dbus-server.cpp)[ 270] main: Discovering devices...
Debug (devicemanager.cpp)[ 358] discover: Starting discovery...
Debug (Configuration.cpp)[ 163] showSetting: Group: (null)
Debug (Configuration.cpp)[ 185] showSetting: vendorid = 4878 (0x0000130E)
Debug (Configuration.cpp)[ 185] showSetting: modelid = 5 (0x00000005)
Debug (Configuration.cpp)[ 209] showSetting: vendorname = Focusrite
Debug (Configuration.cpp)[ 209] showSetting: modelname = Saffire PRO 40

After you see that the device is discovered you can start jack. I don't know if it matters or not, but I created a Preset Name of firewire based on the default.
Interface hw:0 Input/Output channels = 8 each. Note: you may have to setup jack before you do the first step?

Once you start jack you will hopefully see the second green light (indicating fw) light up on the Pro 40. From there it is smooth sailing!
Install audacious and use the jack plugin to test.

And yet all I get is:

00818253238: Fatal (devicemanager.cpp)[ 191] initialize: No firewire adapters (ports) found.
00818253448: Error (ffado-dbus-server.cpp)[ 277] main: Could not initialize device manager
00818254154: Debug (ffado-dbus-server.cpp)[ 202] exitfunction: Debug output flushed...
after running 'ffado-dbus-server &' after a few seconds of turning on the Pro 40.

02:32:43.048 Could not connect to JACK server as client. - Overall operation failed. - Unable to connect to server. Please check the messages window for more info.

If I try to run Jack.

I have tried with Musix 2.0, JackLab, Ubuntu Studio 10.04, Ubuntu 10.04 Netbook Remix with added jackd, ffado and RT-kernel packages compiled from svn, Ubuntu Studio 10.10 changing to old stacks and prior jack and ffado versions, PureDyne 9-11...
This stuff does NOT work guys! In May it will be two years that I am trying to make this happen and/or waiting for DICE support on a release... I wholeheartedly thank the volunteers that pull this project together, but I cannot wait another two years for a DICE support release and FireWire is dying faster as an interface than this project is moving.
I will have to part ways and think of something else. I don't like Macs because they are just doorways to using your credit card for everything new you want to do, so I guess either Windows 7 or a standalone recorder will have to do. If I can ever get all that cash together.
Thank you all again
Good bye.



Does anybody can work with the Saffire PRO 40 in an stable situation? Could you tell me what version of ubuntu,exactly which drivers and which linux kernel? I just got it to work without Realtime mode, and anyway i cannot start from Ardour? Somebody knows why at the Ardour starting you cannot choose "firewire driver" and why in Jack there is a firewire driver option but not a Ffado option?

Is there anybody that could help me pesonally by email? i would appreciate so much!


Try This
1.Do a clean install of Ubuntu 11.04 with updates.

2.Goto "System Settings" then "Users and Groups" then "Advanced Settings" then the "User Privileges" Tab and check the "Use Audio Devices" tab.

3.Goto the "Ubuntu Software Center" and search "ffado" and install the FFADO-Mixer.


5.Turn device on wait a couple of minutes.

6.Start "FFADO MIXER" - It should show your device at this point. (I have noticed that sometimes the mixer doesn't work on some devices.)

7. Start QjackCtl select "firewire' hit start.

I hope this helps.

Yesterday I did a clean install of UbuntuStudio 11.10, 64-bit version with realtime kernel and chose to install only the Audio studio and recording applications. After installation I let the Update Manager run to update the system to the latest versions of all the packages.

In order to have the Saffire Pro 40 work I switched it on BEFORE rebooting the system. After logging in I opened a terminal window and entered:


This produced some warnings and other messages but ended up by showing the available ports on the Saffire.

Next I started Jack Control (included in the Ubuntu Studio distro), selected firewire in the Setup window, and clicked the Start button. Immediately the synchronisation indicator led on the Saffire lit up.

I tested the system with Ardour (also part of UbuntuStudio) and everythig seems to work properly. I even have sound output from my Saffire Pro 40, something I could not get working with UbuntuSutio 11.04. All Xruns, also a problem I had with the previous release of UbuntuStudio, have disappeared as well! In short: it seems I now have a stable system running that fully supports the Focusrite Saffire Pro 40.

Next thing I need to master is the FFado Mixer in order to configure the settings of my Saffire Pro 40.

Care with ffado-mixer.

I experienced many strange behaviours and finally many problems with it during the last days, and presently I will try to reset the Saffire using Windows (I have first to reinstall Windows for the matter).
I am wondering about routing, especially.


I have a Toshiba satellite A300 laptop with "Ricoh Co Ltd R5C592" firewire.

¿Has anyone installed Focusrite saffire pro 40 in Ubuntu/linux with ffado?
¿Is posible to do it?
¿What can I do if it isn't posible?


In the absence of any other feedback I will simply say that I don't know. I don't own either a Ricoh-based interface card or a pro40 so I have no first-hand experience to draw on. We have some information about host controllers at but your particular variant doesn't seem to be listed yet. I would suggest that the best approach in your case is to simply try it and see if it works. If it does, great! Otherwise we'll have to consider alternatives. In either case, report your findings back here so we can add to the general knowledge-base about Ricoh controllers.

Another option may be to subscribe to the ffado-users mailing list and post your question there. You'll reach a different set of people there and among those there may be people who have already tried using your particular chip with ffado.

If it turns out that this controller can't work with audio interfaces, the best approach is probably to obtain a CardBus firewire interface card and use that instead of the built-in interface. Interfaces based on Texas Instruments (TI) chipsets seem to be fairly robust, but as per the host controller page mentioned above there are chips which can work.

I apologise that this information is fairly generic in nature, but without first-hand experience it's a touch tricky to say much more.

Yes I'm using ffado and a Focusrite PRO 40 with my laptop (Asus G2S) wich has a Ricoh based firewire bus. It works OK.
But if you have installed Ubuntu ( Luicd ) you can add TangoStudio or KXStudio repositories, or compile ffado and jack from svn sources.

Hi everybody,

I bought a Saffire Pro 40 and received it yesterday.
I'm awestruck because it was (would have been) working straightforwardly (if I would remember that, at first, I should have to root logging in), at least for the (not so) elementary functions I tested: jack interface, playback and recording. However, I didn't yet have a trial with ffado mixer, I have to learn more about it before.

Note that it was working using the new firewire stack (firewire_ohci, say) included in new kernels
(2.6.35 and 3.0.9, concerning the kernel I tested), with host controller FireWire (IEEE 1394): Ricoh Co Ltd R5C832 IEEE 1394 Controller (rev 05) (as $ lspci returns).
PC is a Dell Precision M6300 laptop, 3 years old running fedora 14 and 16 (with a manually compiled 3.0.2
preempted vanilla kernel for the latter).
Especially, fully preempted 3.0.2 kernel seems to give a very good result (8ms of latency, and probably better if I would be able to solve this damned "fan" (!) dependency of large - occasional - xruns appearance).

So, many, many thanks (and a happy new year, of course) for developpers.
You have done a lot of work, guys !