From: Jano S. <jan...@gm...> - 2013-06-17 17:36:13
|
Hi all, do we know anything about Midas VeniceF? This page http://www.prosoundnetwork.com/article/midas-venicef-analog-mixer-with-firewire/12981says it's built on DICE chipset, so it might be easily supported like the other cards, provided someone finds the vendor & model ID, and maybe routing topology. At least they could be added to device list on the webpage. Jano |
From: Jonathan W. <jw...@ju...> - 2013-06-19 00:31:12
|
Hi Jano > do we know anything about Midas VeniceF? No, not yet. :-) > This page > http://www.prosoundnetwork.com/article/midas-venicef-analog-mixer-with-firewire/12981 > says it's built on DICE chipset, so it might be easily supported like the > other cards, provided someone finds the vendor & model ID, and maybe > routing topology. I believe you are correct. Assuming it is based on DICE, the generic DICE driver and mixer panel should at least do something once we know the vendor and model IDs. Precisely how functional it will be depends on how many vendor-specific extensions they've used, but this is something which is impossible to determine without physically trying it. As you suggested, we will require the vendor/model IDs before FFADO will attempt to drive this device. > At least they could be added to device list on the webpage. Agreed - good idea. See http://www.ffado.org/?q=node/2664. Regards jonathan |
From: Jano S. <jan...@gm...> - 2013-12-12 13:14:17
|
Hi Jonathan, I managed to fix my "home" system, (that's why I was able to report compilation success). I cleaned up everything from /usr/local and also from svn working directory. Then I compiled with PREFIX=/usr and everything was OK after that. I have to do the same on my laptop when time allows. When I can work with my profire on the laptop and find some free time, I'll visit the venice machine, and report. No schedule for now though. Do you have any clue about the clock source thing? There are two leds on the back: Int and Ext. Dice dump shows "Unused, Unused, Internal". When I chose any of the Unused, the led stayed the same. It's just a cosmetic thing, since internal clock is perfectly OK (at least for now)... Jano On Thu, Dec 12, 2013 at 1:08 AM, Jonathan Woithe <jw...@ju...> wrote: > Hi Jano > > On Sun, Dec 08, 2013 at 01:18:29AM +0100, Jano Svitok wrote: > > It seems that my ffado installation is broken somehow - because now I > can't > > stream on any of my two machines. At least one used to work with > > Profire2626 few months ago. > > I guess I have to fix the installation and then venice might work as > well. > > > > Now it's too late, and since I can't get profire to work, there's no > point > > in trying tomorrow. I will get back to this on the next gig :) > > Are you waiting for further feedback here? I have assumed not - I > understand you will come back to it when time allows and follow up as > needed. > > A few minutes ago you posted this in a different thread: > > > However I'm not sure about this, it's just a feeling I got when I was > > compiling ffado on my laptop last week (unsuccesfully :) ) > > I take it that there are still issues on your system. When you get the > chance to post further details we'll do our best to identify the problem. > > > Nevertheless, it is confirmed that it is a DICE device, and we can change > > samplerate, so some progress was made... > > Agreed. Now we've just got to get streaming to work. > > I've had a quick look at the file you made available > (ffado-venice-f32.zip). > This device seems to use two streams for RX and two for TX (see the "DICE > Parameter Space info" reported early in the log). > > The fundamental problem is that the system is dropping packets, and when > faced with this FFADO really isn't going to be able to start reliably. We > therefore need to determine why packets are being dropped on your system. > Could you post the output of ffado-diag if you haven't previously done so? > > Regards > jonathan > |
From: Jonathan W. <jw...@ju...> - 2013-12-12 22:45:09
|
Hi Jano > I managed to fix my "home" system, (that's why I was able to report > compilation success). I cleaned up everything > from /usr/local and also from svn working directory. Then I compiled with > PREFIX=/usr and everything was OK after that. That's good news - I'm pleased to hear it. > I have to do the same on my laptop when time allows. When I can work with > my profire on the laptop and find some free time, > I'll visit the venice machine, and report. No schedule for now though. No problem. > Do you have any clue about the clock source thing? There are two leds on > the back: Int and Ext. Dice dump shows "Unused, Unused, Internal". > When I chose any of the Unused, the led stayed the same. It's just a > cosmetic thing, since internal clock is perfectly OK (at least for now)... It's a bit hard to know without having the Venice documentation (and even then, the documentation might not explain it). I can think of a few possibilities. 1) The "Ext" LED has been provided for future expansion and is currently unused. 2) The clock selection of the Venice is automatic, with an external source selected whenever a digital input is connected. This of course assumes that the Venice has some form of digital input. 3) The "unused" reports from the Dice dump are interesting. Either the internal source is the only one available for now (see also point 1) or the strings change dynamically if a valid external clock source is detected on the system. 4) It's very unlikely, but perhaps the Venice does clock selection in a vendor-specific, non-standard way. Some experimentation may be needed to determine what the situation is. Regards jonathan |
From: Jörn N. <net...@st...> - 2013-06-20 10:24:47
|
On 06/19/2013 02:30 AM, Jonathan Woithe wrote: > Hi Jano > >> do we know anything about Midas VeniceF? > > No, not yet. :-) > >> This page >> http://www.prosoundnetwork.com/article/midas-venicef-analog-mixer-with-firewire/12981 >> says it's built on DICE chipset, so it might be easily supported like the >> other cards, provided someone finds the vendor & model ID, and maybe >> routing topology. > > I believe you are correct. Assuming it is based on DICE, the generic DICE > driver and mixer panel should at least do something once we know the vendor > and model IDs. Precisely how functional it will be depends on how many > vendor-specific extensions they've used, but this is something which is > impossible to determine without physically trying it. > > As you suggested, we will require the vendor/model IDs before FFADO will > attempt to drive this device. > >> At least they could be added to device list on the webpage. > > Agreed - good idea. See http://www.ffado.org/?q=node/2664. fwiw, i will be mixing a show on such a midas frame by the end of the month - if everything runs smoothly, i'll try and obtain some ffado-diag data and see if i can get it to stream. best, jörn -- Jörn Nettingsmeier Lortzingstr. 11, 45128 Essen, Tel. +49 177 7937487 Meister für Veranstaltungstechnik (Bühne/Studio) Tonmeister VDT http://stackingdwarves.net |
From: Jonathan W. <jw...@ju...> - 2013-06-20 10:31:07
|
Hi Jörn On Thu, Jun 20, 2013 at 12:24:26PM +0200, Jörn Nettingsmeier wrote: > >>do we know anything about Midas VeniceF? > : > fwiw, i will be mixing a show on such a midas frame by the end of > the month - if everything runs smoothly, i'll try and obtain some > ffado-diag data and see if i can get it to stream. That sounds like a plan. In order to get streaming happening it will be necessary to add the device's vendor and model IDs into the FFADO file "configuration" and indicate that the drive is 20 (DICE). One easy way to get the required IDs is to run gscanbus. If you click on the icon corresponding to the mixer you should get a text window with the required information in it. Even though we currently need only the vendor and model IDs, it would be good to post all the information returned by gscanbus for future reference. Regards jonathan |
From: Stefan R. <st...@s5...> - 2013-06-22 08:10:05
|
On Jun 20 Jonathan Woithe wrote: > In order to get streaming happening it will be necessary to add the device's > vendor and model IDs into the FFADO file "configuration" and indicate that > the drive is 20 (DICE). One easy way to get the required IDs is to run > gscanbus. If you click on the icon corresponding to the mixer you should > get a text window with the required information in it. Even though we > currently need only the vendor and model IDs, it would be good to post all > the information returned by gscanbus for future reference. If gscanbus is not provided by the distribution of choice, the command "grep -r . /sys/bus/firewire/devices/*" is an alternative to collect the device identifiers. -- Stefan Richter -=====-===-= -==- =-==- http://arcgraph.de/sr/ |
From: Jano S. <jan...@gm...> - 2013-12-06 22:22:11
|
Hi all, I've had possibility to do some tests with Midas Venice F32. It seems that it is DICE indeed. I have put some information on wiki ( http://subversion.ffado.org/wiki/Midas). If I read the output of ffado-test Discover correctly, then: - the card provides 32 inputs and 32 outputs (the channel names seem to be quite right) - clock source internal and external (from manual, I don't understand the output) - sample rate 44.1k or 48k I did an attempt at configuration entry before running ffado-test, so there is sticking line "mixer: Generic_Dice_EAP". I'm not sure if model in configuration should be 0x1 or 0x4029bc. I suppose the configuration entry should look like this (without mixer line): I don't know if F16 and F24 have the same ID's or not (I checked only F32). { # MIDAS Venice F32 vendorid = 0x0010C73F modelid = 0x00000001; vendorname = "Midas"; modelname = "Venice F32"; driver = "DICE"; } I could not check streaming as I didn't have jack/ardour installed on the machine. I will try tomorrow. I will have access to the console this weekend and some time to play. What tests should I run? Jano On Sat, Jun 22, 2013 at 10:09 AM, Stefan Richter <st...@s5...>wrote: > On Jun 20 Jonathan Woithe wrote: > > In order to get streaming happening it will be necessary to add the > device's > > vendor and model IDs into the FFADO file "configuration" and indicate > that > > the drive is 20 (DICE). One easy way to get the required IDs is to run > > gscanbus. If you click on the icon corresponding to the mixer you should > > get a text window with the required information in it. Even though we > > currently need only the vendor and model IDs, it would be good to post > all > > the information returned by gscanbus for future reference. > > If gscanbus is not provided by the distribution of choice, the command > "grep -r . /sys/bus/firewire/devices/*" is an alternative to collect the > device identifiers. > -- > Stefan Richter > -=====-===-= -==- =-==- > http://arcgraph.de/sr/ > |
From: Jonathan W. <jw...@ju...> - 2013-12-07 03:51:23
|
Hi Jano [CC list trimmed] > I've had possibility to do some tests with Midas Venice F32. It seems that > it is DICE indeed. I have put some information on wiki ( > http://subversion.ffado.org/wiki/Midas). Great - thanks for the information. > If I read the output of ffado-test Discover correctly, then: > > - the card provides 32 inputs and 32 outputs (the channel names seem to be > quite right) > - clock source internal and external (from manual, I don't understand the > output) > - sample rate 44.1k or 48k That would be consistent with a majority of mixers of this nature. > I did an attempt at configuration entry before running ffado-test, so there > is sticking line "mixer: Generic_Dice_EAP". > > I'm not sure if model in configuration should be 0x1 or 0x4029bc. Based on the information parsed from the configrom which you put on the wiki, the correct value for the modelid should be 0x000001. > I suppose the configuration entry should look like this (without mixer > line): I don't know if F16 and F24 have the same ID's or not (I checked > only F32). > > { # MIDAS Venice F32 > vendorid = 0x0010C73F > modelid = 0x00000001; > vendorname = "Midas"; > modelname = "Venice F32"; > driver = "DICE"; > } This seems consistent with what the parsed configrom says. I have added this to our development trunk - see r2462. I've included the generic DICE mixer for the moment since it may provide something useful. > I could not check streaming as I didn't have jack/ardour installed on the > machine. I will try tomorrow. Sure, no worries. > I will have access to the console this weekend and some time to play. What > tests should I run? There are probably two things to try. The first would be to run ffado-mixer and see what controls it provides (and whether any of them make sense). The second thing is to see if jackd/ffado starts up with this device. You could try starting jackd on a command line if it's more convenient: jackd -P70 -R -dfirewire -p512 -n3 If jackd holds up and doesn't crash then it's likely that FFADO is successfully streaming to and from the device. At this point you could try playing audio using ecasound or some other jack-enabled program. If on the other hand jackd crashes it means we have a problem - perhaps rerun it using jackd -P70 -R -dfirewire -p512 -n3 -v6 >& /tmp/ffado-venice_f32.log and make the ffado-venice_f32.log for us to have a look through. Regards jonathan |
From: Jano S. <jan...@gm...> - 2013-12-07 23:43:00
|
Hi Jonathan, I did as you told me. I could switch sample rate from 44.1k to 48k and back (confirmed by LEDs on the console). I could not change clock source (neither of the two unused changed anything on the console). I tried to start jack, but it stopped right away. The log is at http://zvukari.sk/ffado/ffado-venice-f32.zip (~200 kB, I suppose it's too long for the list). I will recompile ffado and try again. I will check my installation again my (known to be working) profire2626. Jano |
From: Jano S. <jan...@gm...> - 2013-12-08 00:18:36
|
It seems that my ffado installation is broken somehow - because now I can't stream on any of my two machines. At least one used to work with Profire2626 few months ago. I guess I have to fix the installation and then venice might work as well. Now it's too late, and since I can't get profire to work, there's no point in trying tomorrow. I will get back to this on the next gig :) BUT, if you spot something obvious, I will gladly try anything withtin those few ours I can devote to this tomorrow. Nevertheless, it is confirmed that it is a DICE device, and we can change samplerate, so some progress was made... Jano On Sun, Dec 8, 2013 at 12:42 AM, Jano Svitok <jan...@gm...> wrote: > > Hi Jonathan, > > I did as you told me. I could switch sample rate from 44.1k to 48k and > back (confirmed by LEDs on the console). I could not change clock source > (neither of the two unused changed anything on the console). > > I tried to start jack, but it stopped right away. The log is at > http://zvukari.sk/ffado/ffado-venice-f32.zip (~200 kB, I suppose it's too > long for the list). > > I will recompile ffado and try again. I will check my installation again > my (known to be working) profire2626. > > Jano > |
From: Jonathan W. <jw...@ju...> - 2013-12-12 00:09:22
|
Hi Jano On Sun, Dec 08, 2013 at 01:18:29AM +0100, Jano Svitok wrote: > It seems that my ffado installation is broken somehow - because now I can't > stream on any of my two machines. At least one used to work with > Profire2626 few months ago. > I guess I have to fix the installation and then venice might work as well. > > Now it's too late, and since I can't get profire to work, there's no point > in trying tomorrow. I will get back to this on the next gig :) Are you waiting for further feedback here? I have assumed not - I understand you will come back to it when time allows and follow up as needed. A few minutes ago you posted this in a different thread: > However I'm not sure about this, it's just a feeling I got when I was > compiling ffado on my laptop last week (unsuccesfully :) ) I take it that there are still issues on your system. When you get the chance to post further details we'll do our best to identify the problem. > Nevertheless, it is confirmed that it is a DICE device, and we can change > samplerate, so some progress was made... Agreed. Now we've just got to get streaming to work. I've had a quick look at the file you made available (ffado-venice-f32.zip). This device seems to use two streams for RX and two for TX (see the "DICE Parameter Space info" reported early in the log). The fundamental problem is that the system is dropping packets, and when faced with this FFADO really isn't going to be able to start reliably. We therefore need to determine why packets are being dropped on your system. Could you post the output of ffado-diag if you haven't previously done so? Regards jonathan |