Moseph
Max Output Level: -89 dBFS
- Total Posts : 85
- Joined: 2007/06/13 17:31:40
- Status: offline
New Interface Time
Okay, so after more than 18 months of trying, I can't get suitably stable operation with my Firestudio 2626 at suitably low latencies with Sonar. I've done the DPC checks, and I've tweaked my XP system pretty well (so far as I know). All software/firmware/drivers are at the most recent release. Sonar just flat-out doesn't want to work well at low latencies. Cross-checked with Cubase LE 1.07 and it works better, but still struggles a little bit at the 128-sample block sizes and below. Let me elaborate: I get good takes/captures, but you don't get that impression during the real-time fold-back. It sounds like there's pops/drops all over the place, even though there's not. I can't have monitoring that makes me think there's problems when there are not, and more importantly, I can't record somebody else if they can't reliably monitor themselves during the take. So that means I need to bite the bullet and go get myself a different interface. I don't want to downgrade (I still have my old Tascam FW-1804 if I was willing to do that). That means I need to match (or beat) the core functionality of the Firestudio: - at least 26 inputs (including 2 I/O via S/PDIF): this one is the important one - at least 8 outputs, preferably with options for direct/CPU outputs - 8+ analog inputs with preamps/phantom power - at least 1 MIDI in and 1 MIDI out - 1-2U Rack-mount units - Firewire 400 (my external HDD's are all USB 2.0) It's also gotta be stable using Sonar 8.5, and preferably whatever else wants to access it. Pro Tools compatibility is something I've considered, but is not really a priority. The budget is somewhere in the range of $1000 all told (I've been saving for this day for awhile hoping I wouldn't need to spend the money). I've done some research on this, and have come up with some ideas. MOTU 828mk3 This one is probably my top choice for the moment. Pros are that it has most of the features I'm looking for, MOTU's reputation for reliability, and single rack-space. Cons are the lack of preamps, meaning I'd probably need to also get another rackmount unit with 8 pres to expand this thing. That's still within the budget, but is pretty close to maxing it out. MOTU Travelermk3 I'm less keen about this one: it's got the same functionality as the 828, plus 2 preamps and the option for bus-powering off firewire. Cons are that it only gives me 4 preamps, so I'll want to add 4 more anyway (which will likely be the same as adding 8 in terms of cost), also it's got a weird chassis design (with power, firewire and MIDI on the side, rather than front/rear): that's not something I like when I'm going to keep it racked up in a box. Also it's about $100 more expensive than the 828. M-Audio ProFire 2626 Pros include a moderate price compared to the above 2, plus it lets me do Pro Tools M-Powered with full I/O, which would be nice, but isn't something I'm too concerned with. Cons include the connector layouts all on the back (I really like that I can plug straight into the front of the Firestudio), and JetPLL, which I'm pretty sure means DICE II chipsets. DICE II is the probably issue with the Firestudio, so I might just be a frying pan/fire situation. I'm also not super-keen on the need for a breakout cable for MIDI, clock, or S/PDIF. M-Audio ProFire Lightbridge This one is interesting to me. Pros include a potential higher channel count, Pro Tools M-Powered again, and a nice price tag. Cons are the breakout cable again, and also the complete lack of preamps. This actually isn't a huge deal since I've already got 16 channels of ADAT lightpipe pres (actually 24, but I don't want to revert back to the SMPro Unit and would rather get rid of it), so in all I could get back up to my goal of 26 for less than the cost of the other units here. However, I'm not sure about DICE II in this thing, so that's something else to consider. Focusrite Saffire Pro 26 Pros are that it has all the basic features I'm looking for. Cons are that it's not made anymore, so if I can find one, I'm buying used (though this will probably help with price). The newer Saffire Pro series aren't really an option here since they're using DICE II, and they're also much more expensive (I'd need the 56 to reach my I/O goals), and they also have absolutely stupid design (snakes don't always have fan-out to go to the front and back to plug in). Also, I've dealt with the 56 and I wasn't particularly impressed with the pres (particularly those preamp emulations: that's a feature I'm paying for but will never use). Did I miss anything pertinent? Does anyone have info on non-Presonus DICE II devices and Sonar behaving more reliably? Anybody have any strong opinions one way or the other about an interface they happen to have or have used? Does anybody have a success/failure story with any of the above interfaces? Advice is definitely welcome. Thanks.
post edited by Moseph - 2010/10/18 09:20:05
|
daveny5
Max Output Level: 0 dBFS
- Total Posts : 16934
- Joined: 2003/11/06 09:54:36
- Location: North Carolina
- Status: offline
Re:New Interface Time
2010/10/18 10:11:44
(permalink)
What do you call low latency? What latency are you getting? I get 5.8msec which is fine. Some people get lower than that. With Firewire devices, the most important component is the firewire port on your computer. If it doesn't use the TI (Texas Instruments) chipset, you will have problems. That has been documented in this forum hundreds of times.
Dave Computer: Intel i7, ASROCK H170M, 16GB/5TB+, Windows 10 Pro 64-bit, Sonar Platinum, TASCAM US-16x08, Cakewalk UM-3G MIDI I/F Instruments: SL-880 Keyboard controller, Korg 05R/W, Korg N1R, KORG Wavestation EX Axes: Fender Stratocaster, Line6 Variax 300, Ovation Acoustic, Takamine Nylon Acoustic, Behringer GX212 amp, Shure SM-58 mic, Rode NT1 condenser mic. Outboard: Mackie 1402-VLZ mixer, TC Helicon VoiceLive 2, Digitech Vocalist WS EX, PODXTLive, various stompboxes and stuff. Controllers: Korg nanoKONTROL, Wacom Bamboo Touchpad
|
AT
Max Output Level: 0 dBFS
- Total Posts : 10654
- Joined: 2004/01/09 10:42:46
- Location: TeXaS
- Status: offline
Re:New Interface Time
2010/10/18 10:51:10
(permalink)
You are asking a lot for $1000. Most of the candidates use the Dice II chip at that price, and you say you are having a problem already w/ a dice II chip. I do know that TC has just released new drivers (2.5) for their interfaces and it is supposed to be very low latency now (you can check their user forums for info and users comments). You could cobble together something from them if it works - I haven't updated my TCK 48 yet. But a TCK 48 & twin ought to reach your specs (you can stack the units). Great sound, great built in DSP, good preamps and a nice remote. Like daveny I run a dual core at about 6 ms delay w/o the new drivers. Another great feature is 0 delay mix, which is what you seem to have a problem w/. I"m not sure how a2nd unit works with the TC mixer, but you can send the direct ins to 2 subs (w/ effects) and the main w/ no disernable delay for monitoring. Most other systems would be more expensive - the SSL is 24 i/o but w/ the pci card close to $3000. Other systems include hard disk hardware units, tho most top out @ 24 channels. The m-audio and Presonus ADAT to FW would also work, but I've never used them. But I would think if you are going for low latency and recording so many tracks at once you might have the same problem, since you have to stream through SONAR and back out to get effects/mixing. @
https://soundcloud.com/a-pleasure-dome http://www.bnoir-film.com/ there came forth little children out of the city, and mocked him, and said unto him, Go up, thou bald head; go up, thou bald head. 24 And he turned back, and looked on them, and cursed them in the name of the Lord. And there came forth two she bears out of the wood, and tare forty and two children of them.
|
Moseph
Max Output Level: -89 dBFS
- Total Posts : 85
- Joined: 2007/06/13 17:31:40
- Status: offline
Re:New Interface Time
2010/10/18 10:53:54
(permalink)
daveny5 What do you call low latency? What latency are you getting? I get 5.8msec which is fine. Some people get lower than that. With Firewire devices, the most important component is the firewire port on your computer. If it doesn't use the TI (Texas Instruments) chipset, you will have problems. That has been documented in this forum hundreds of times. In Cubase I can use 256 sample blocks pretty reasonably. At 44.1kHz this means 5.805 msec, at 48 kHz this means 5.3 msec. This is still acceptable. The next step up in the driver is 512 samples, so that means 11.6 msec and 10.6 msec respectively. That's marginal: most folks are fine, but I have a few people I work with who complain. One more step up is 1024, which is on the order of 20 msec and most of the musicians I work with can't deal with that sort of latency (most importantly, I am one of them). I generally work at 44.1 kHz, but want to maintain the possibility of 48 kHz in case I ever end up trying my hand at audio for video projects (I probably will in the future). Sonar can, for the most part, match the behavior on the front-end. The real problem is that the monitoring is not stable: it sounds like there's lots of dropouts and problems in the playback, which makes playing along to the monitoring very difficult. This is not a huge problem for some projects where playback is not needed, but it absolutely kills the potential to overdub (most of my projects will use overdubs somewhere). Sonar also has a tendency to dropout after a few minutes at the 256-sample block size. Cubase doesn't do this, so I believe my problems have to do with the interaction between the Firestudio and Sonar, and not just either part of the equation individually. System-wise, I believe I've done all the necessary tweaks to Windows XP (SP3, fully updated) that might interfere with performance. This behavior is identical in two different PCe Firewire cards: a SIIG card using the VIA chipset, and a StarTech EC13942 with a TI chipset. Both these chipsets are recommended by Presonus, and the StarTech card is explicitly listed as 1 of 2 Express cards recommended by Presonus. I'm using integrated Intel video chips, so there cannot be any issues with ATI RADEON or nForce4 compatibility. DPC analyzer puts me on the order of 180 microseconds max delay when I disable Wi-Fi, Ethernet, and ACPI Battery Control Method. There are no IRQ conflicts with any hardware. All external hard drives run off USB 2.0, by themselves on a free port and spin at 7200 rpm. Anti-virus is disabled during audio work. The internal drive is only running software, and while it's only spinning at 5400 rpm, I'm not inclined to believe it's the cause of the problem since it's not doing any of the heavy lifting with respect to audio streaming. Processor is a Core Duo 2 1.8 GHz, 3 GB RAM. CPU metering shows no spikes, jumps or abnormal behavior during recording, and both CPU and RAM usage is below 10% during core tests. No plugins are in use during the core tests. The most recent test was with 12 channels across all inputs of the Firestudio (i.e., the analog connections, both banks of ADAT Lightpipe inputs, and also the S/PDIF input). Also important to note is that I tested this on my old laptop as well, which was rock-solid using Sonar 7 and a Tascam FW-1804. Similar problems exist when I switched over to the Firestudio. In all situations, I'm using ASIO drivers and not WDM drivers. Testing WDM drivers didn't provide any benefit in Sonar, and also had the added disadvantage of limiting me to 8 channels of I/O, which is very much the opposite of what I'm looking to do (I can only guess what brain trust at Presonus decided this was a good implementation to support for a 26-channel interface like the Firestudio). If you need any more tech data, let me know and I'll see what I can do.
post edited by Moseph - 2010/10/18 11:09:54
|
Moseph
Max Output Level: -89 dBFS
- Total Posts : 85
- Joined: 2007/06/13 17:31:40
- Status: offline
Re:New Interface Time
2010/10/18 10:59:31
(permalink)
AT You are asking a lot for $1000. Most of the candidates use the Dice II chip at that price, and you say you are having a problem already w/ a dice II chip. I do know that TC has just released new drivers (2.5) for their interfaces and it is supposed to be very low latency now (you can check their user forums for info and users comments). You could cobble together something from them if it works - I haven't updated my TCK 48 yet. But a TCK 48 & twin ought to reach your specs (you can stack the units). Great sound, great built in DSP, good preamps and a nice remote. Like daveny I run a dual core at about 6 ms delay w/o the new drivers. Another great feature is 0 delay mix, which is what you seem to have a problem w/. I"m not sure how a2nd unit works with the TC mixer, but you can send the direct ins to 2 subs (w/ effects) and the main w/ no disernable delay for monitoring. Most other systems would be more expensive - the SSL is 24 i/o but w/ the pci card close to $3000. Other systems include hard disk hardware units, tho most top out @ 24 channels. The m-audio and Presonus ADAT to FW would also work, but I've never used them. But I would think if you are going for low latency and recording so many tracks at once you might have the same problem, since you have to stream through SONAR and back out to get effects/mixing. @ The SSL is way out of my budget. We're also talking fire-wire because I'm dealing with a PCe cardport on a laptop. If I was going to blow the budget, it'd be on a RME Fireface 800, but that's a little more than I'm comfortable spending all at once. Replacing the computer is a last-ditch effort, since I bought about 5 months after the Firestudio, it's still a relatively clean install (I Firefox, added OpenOffice.org, anti-virus, and other audio software only: none of which is used while Sonar is open), and so far as my tests go it's still malware free. I don't know that $1000 is really that far a reach for this, although it's rapidly looking that way as more devs move to DICE II. However, having read these forums and some others, it looks like the real problem manufacturers with DICE II units are Presonus and TC Electronic (which kind of blows since they make my 1st and 2nd choice hardware configurations respectively). If anybody's got success stories where they started with issues on a DICE II unit and eventually resolved them, I'm more than interested.
|
AT
Max Output Level: 0 dBFS
- Total Posts : 10654
- Joined: 2004/01/09 10:42:46
- Location: TeXaS
- Status: offline
Re:New Interface Time
2010/10/18 13:50:02
(permalink)
As I wrote above, TC has new drivers out. The previous iteration got me down to 5 ms latency or 11 round trip. Several users on the Forum have halved their latency using the new drivers. TC developed DICE 2 and a bunch of companies jumped on it since theoritically it would do exactly what you want to do. Several years later and it seems to finally be reaching its potential. As I also said, I haven't upgraded to the drivers, but when I do I expect it to help. Most of the horror stories w/ DICE II are from the first year and they seem to have worked the major problems out. For TC itself, they do both Mac and PC so they have to work on both. On my system the drivers are stable and work well - nothing but less than stellar latency. The TC Konnekt 48 only has 22 inputs - 12 lines (4 preamps), SPdif and 8 ADAT in. The ADAT includes SMUX support, so 8 track at 96 kHz but not 16 using both ADAT i/os. The sound is very good imo - on par w/ RME (they use the convertor chip and the electronics around it must be pretty good). Again, as above, I'm not really sure what is causing your dropouts, crackles for monitoring. It it is band at once direct monitoring (like I described from my TCK 48) would allow you to set up monitoring mixes for the band. If it is happening on overdubbing, it could be a CPU problem. At low latencies and a lot of mixing effects thrown on top you will get flaws - but you should hear them when just playing back a track. If you only hear them when recording, you are simply too close to the maximum CPU usuage. Again, something like the TC w/ built in DSP can help, or saving a low CPU, effect lite version of the song to play back for overdubbing can solve the problem. So maybe you don't need a new interface or computer. Just an idea. @
https://soundcloud.com/a-pleasure-dome http://www.bnoir-film.com/ there came forth little children out of the city, and mocked him, and said unto him, Go up, thou bald head; go up, thou bald head. 24 And he turned back, and looked on them, and cursed them in the name of the Lord. And there came forth two she bears out of the wood, and tare forty and two children of them.
|
Moseph
Max Output Level: -89 dBFS
- Total Posts : 85
- Joined: 2007/06/13 17:31:40
- Status: offline
Re:New Interface Time
2010/10/18 14:27:33
(permalink)
AT Again, as above, I'm not really sure what is causing your dropouts, crackles for monitoring. It it is band at once direct monitoring (like I described from my TCK 48) would allow you to set up monitoring mixes for the band. If it is happening on overdubbing, it could be a CPU problem. At low latencies and a lot of mixing effects thrown on top you will get flaws - but you should hear them when just playing back a track. If you only hear them when recording, you are simply too close to the maximum CPU usuage. Again, something like the TC w/ built in DSP can help, or saving a low CPU, effect lite version of the song to play back for overdubbing can solve the problem. So maybe you don't need a new interface or computer. Just an idea. @ Sorry, guess this wasn't clear from the previous posts: the bad monitoring happens without any plugins in operation. None are even loaded up. It's just the pass-through from the input to the output. This is true of the overdubs I was talking about as well (though overdubbing 2 channels onto an existing set of 12+ channels doesn't have the same issues). I generally don't have the issue while just playing back. Actually, just monitoring without recording doesn't really have any issues either. It's definitely something weird, particularly since I've done a couple of sessions working through it now and the inputs are, for the most part, successfully captured despite the monitoring issues. Both Windows and Sonar monitoring show very low usage of CPU cycles and RAM, with no apparent spiking at any point. Generally less than 10% usage. Like I said, I suspect there's just something about the interaction between Sonar and the Firestudio drivers that is causing this issue. It may be specific to my system(s) specifically, but since other DAW software is more stable on both computers, I'm betting that's not the case. Thanks for the heads up on the TC Konnekt 48's I/O as well. I didn't realize that the ADAT Lightpipe's S/MUX channels were S/MUX-only. Unfortunately, that likely a deal-breaker for me, as the TC unit is just too expensive for me to justify the downgrade in I/O.
post edited by Moseph - 2010/10/18 14:29:21
|
AT
Max Output Level: 0 dBFS
- Total Posts : 10654
- Joined: 2004/01/09 10:42:46
- Location: TeXaS
- Status: offline
Re:New Interface Time
2010/10/18 17:58:17
(permalink)
That is weird, esp. as other software doesn't exhibit the problem. Not a CPU problem. Hmmm. One of those things - probably Presonus. Yea, I would check your hopefull interface to make sure they aren't double counting inputs. You know those marketing guys. best of luck with a new interface and let us know how it works. @
https://soundcloud.com/a-pleasure-dome http://www.bnoir-film.com/ there came forth little children out of the city, and mocked him, and said unto him, Go up, thou bald head; go up, thou bald head. 24 And he turned back, and looked on them, and cursed them in the name of the Lord. And there came forth two she bears out of the wood, and tare forty and two children of them.
|
bvideo
Max Output Level: -58 dBFS
- Total Posts : 1707
- Joined: 2006/09/02 22:20:02
- Status: offline
Re:New Interface Time
2010/10/18 21:36:35
(permalink)
Moseph ... In Cubase I can use 256 sample blocks pretty reasonably. At 44.1kHz this means 5.805 msec, at 48 kHz this means 5.3 msec. This is still acceptable. The next step up in the driver is 512 samples, so that means 11.6 msec and 10.6 msec respectively. That's marginal: most folks are fine, but I have a few people I work with who complain. One more step up is 1024, which is on the order of 20 msec and most of the musicians I work with can't deal with that sort of latency (most importantly, I am one of them). I generally work at 44.1 kHz, but want to maintain the possibility of 48 kHz in case I ever end up trying my hand at audio for video projects (I probably will in the future). ... Depending on how new the driver is, there are some versions of the DICE II driver that allow you to type in the desired buffer size. In that case you could try 384 to aim for 8.7 / 8.0, if that's enough improvement. Bill B.
|
Moseph
Max Output Level: -89 dBFS
- Total Posts : 85
- Joined: 2007/06/13 17:31:40
- Status: offline
Re:New Interface Time
2011/01/12 19:48:44
(permalink)
Okay, I ended up going with the MOTU 828mk3 a little while ago, and I've done a little bit of testing and so far things seem solid again. At this point I'm pretty confident that the problem was largely to do with the interaction between Sonar and the Firestudio. Recently did a "small load" stress testing without issues: 13 simultaneous inputs at 5msec latency for about 25 minutes and no issues whatsoever. Hopefully my troubles are over for now. Thanks for the all the help everybody.
|
IronSound
Max Output Level: -89 dBFS
- Total Posts : 61
- Joined: 2010/05/06 19:47:16
- Location: Lincoln. NE
- Status: offline
Re:New Interface Time
2011/01/13 02:03:01
(permalink)
Hello... I found with my Presonus Projects, that shutting off Disc Caching in Sonar and/or Windows in (hard drive properties)... made all my pops/clicks/drops go away... give those a try.
post edited by IronSound - 2011/01/13 02:05:25
|