Great Combo - BFM DSR280s & SAC

Post your reviews and pictures here.
Message
Author
User avatar
BrentEvans
Posts: 3041
Joined: Thu Oct 09, 2008 10:38 am
Location: Salisbury, NC

Re: Great Combo - BFM DSR280s & SAC

#16 Post by BrentEvans »

UROK wrote: How about SAC for recorded music, club and DJ applications? Or even a mix of recorded and live sound, say a scratch artist with a live bassist, for example? (I am building DR280s and asking about SAC so hope this isn't a threadjack!) [If so...Mr Moderator. Sorry.]
SAC takes the place of (up to) a 72 channel mixing console and all the processing to go along with it. It will mix and process your audio quite well, but it isn't a DJ mixer, and won't perform certain functions well (like crossfading) without a hardware fader pack. None of the supported hardware fader options are the DJ type.

Now if your DJ is sending you a stereo feed, and you want to mix him with something else... it'll do that... but the best fit is to use it for what it was designed - a large live console.
99% of the time, things that aren't already being done aren't being done because they don't work. The other 1% is split evenly between fools and geniuses.

UROK

Re: Great Combo - BFM DSR280s & SAC

#17 Post by UROK »

Thanks Brent.

Yeah, sorry, it was more for controlling the main out stereo feed I was thinking and putting it together with other live signals. Frightening to think what a "red line monkey" DJ would end up doing if they were given access to SAC!

But my impression was it was for those large console type situations.

I can't see it being justifiable on a cost basis for club/DJ work.

If I ever move into live sound though...(live sound here is lamentable with guys forking out on huge systems and thinking they just plug 'em in and turn it up and you get concert sound...such as with the Modernsky music festival next week where they've flown in Boz from LivePA in NYC to sort out their Eventide system they can't get sounding right...)

el_ingeniero
Posts: 931
Joined: Sun Jul 18, 2010 11:46 pm
Location: Minneapolis, MN

Re: Great Combo - BFM DSR280s & SAC

#18 Post by el_ingeniero »

SAC has it's place. If you need a 32+ channel digital console on a 16 channel digital budget, it's your huckleberry.

The cons:

1) Very rider unfriendly. Although, some acts even don't like a Yamaha LS9.

2) Having it run on a general purpose computer means it's never going to be quite as reliable as a console: too many things that can break (although the new SSDs help a lot). I'd want a backup installation of some kind.

3) And it runs on Windows (bleah), although to be honest, some digital consoles run Windows too. But that's more of a personal bias of mine.

4) It needs to be able to use multiple processors, badly: there is a top limit on how fast individual processors run, and processor manufacturers hardly make general purpose CPUs with one processor anymore because of that. But that's not going to happen: SAC is a one-man show that's not up to the development effort.

All that being said, if I ran a small sound company looking to get across the gap between doing local bands and groups with a rider, I'd use it until I could get my hands on a rider friendly digital console, and keep it using it as long as I am doing local bands. But that's just me.

User avatar
BrentEvans
Posts: 3041
Joined: Thu Oct 09, 2008 10:38 am
Location: Salisbury, NC

Re: Great Combo - BFM DSR280s & SAC

#19 Post by BrentEvans »

In all fairness, some of these points are a matter of perspective.
el_ingeniero wrote: 1) Very rider unfriendly. Although, some acts even don't like a Yamaha LS9.
This is true. If you're a good salesman, you might be able to get past this with some... and another solution is to always supply a board op. This way, if a GE needs to mix, they can "produce" and you or your person run the system.
2) Having it run on a general purpose computer means it's never going to be quite as reliable as a console: too many things that can break (although the new SSDs help a lot). I'd want a backup installation of some kind.

3) And it runs on Windows (bleah), although to be honest, some digital consoles run Windows too. But that's more of a personal bias of mine.
Windows is quite stable... until you start gomming it up with all manner of crum. A SAC PC should be dedicated and free of other gom and crum. I limit mine to SAC, Reaper, Winamp, and Fomine Net Send GUI. I do a stripped down windows install, no games, no screen savers, etc. It's stable 24/7. No unnanounced crashes in over a year... the last one was due to a lightning strike.
4) It needs to be able to use multiple processors, badly: there is a top limit on how fast individual processors run, and processor manufacturers hardly make general purpose CPUs with one processor anymore because of that. But that's not going to happen: SAC is a one-man show that's not up to the development effort.
SAC is multiple processor aware. Right now, the Intel platform allows any core in a multicore system instant access to RAM. This is what actually causes the glitching with multicore machines... but the menu option for forcing the SAC audio engine to one CPU is very effective at preventing this, then the other CPU can handle Video, MIDI, etc.

In any case, the fastest dual core machines around today are more than capable of running a very robust 72 channel mix with a bunch of monitor mixes, and crazy amounts of processing with room to spare. I can run 48 channels and 10 monitor mixes with multiple verbs, comps, eqs, gates, VST plugs, etc.... and on my E5200 system (not the fastest in the world) I top out at about 30-40%. The system is easily stable at double that load. Processor speed is no simply not an issue these days.

As for development... the development is astounding for a one man business. The developer's current focus is expanding the feature set and making the application more powerful. He always comes p with good stuff.
All that being said, if I ran a small sound company looking to get across the gap between doing local bands and groups with a rider, I'd use it until I could get my hands on a rider friendly digital console, and keep it using it as long as I am doing local bands. But that's just me.
Ultimately you do what makes you money. If that means owning a rider-friendly console, you do it. However... the new Midas stage boxes connect to ASIO drivers on a PC through a standard ethernet port... they're relatively affordable, and they're "MIDAS". This is easy to work with from a sales perspective... "We us a customized software based digital mixer with a Midas front end.. this allows us to give you XX channels and 24 discrete monitor mixes, etc, etc... Things like that are going to make SAC more and more acceptable... it's just a matter of explanation in some cases. Unless you're playing in the "super-pro" field where the GEs want to just load up the same console off a USB stick and go to town, SAC still has a place in the game.
99% of the time, things that aren't already being done aren't being done because they don't work. The other 1% is split evenly between fools and geniuses.

UROK

Re: Great Combo - BFM DSR280s & SAC

#20 Post by UROK »

Wow!

And there it is in two posts! Why this forum is unparalleled.

Amazing knowledge. Respectfully put. I have learned heaps.

Thank you for sharing.

Dunno what 'gom' and 'crum' are but then again I haven't watched Star Trek in a while. :mrgreen:

Thanks again. You guys are awesome. :clap:

User avatar
BrentEvans
Posts: 3041
Joined: Thu Oct 09, 2008 10:38 am
Location: Salisbury, NC

Re: Great Combo - BFM DSR280s & SAC

#21 Post by BrentEvans »

UROK wrote:Wow!

And there it is in two posts! Why this forum is unparalleled.

Amazing knowledge. Respectfully put. I have learned heaps.


Thank you for sharing.
That's what it's all about
Dunno what 'gom' and 'crum' are but then again I haven't watched Star Trek in a while. :mrgreen:
I don't suppose a Brit living in China would know about gom and crum. Allow me to define:

gom (n). [gahm]. An unidentified mass of substance or substances, often which have or have had some actual meaningful purpose, but are installed in a haphazard manner, or with simple carelessness, resulting in a royal mess. He looked behind his friend's TV and found a gom of cables and wires.

gom (v). [gahm] The process by which gom (n) occurs.

crum (n) [kruhm] An unidentified mass, substance, or general object which has found its way into something useful, resulting in a loss of performance. Often crum occurs because gom is in the way. He had all kinds of crum on his computer which really slowed it down.

Hope that helps.
Thanks again. You guys are awesome. :clap:
We'll be here all week.

Image

BTW, @ Chris, you should really steal that smiley. It's awesome.
99% of the time, things that aren't already being done aren't being done because they don't work. The other 1% is split evenly between fools and geniuses.

SirNickity
Posts: 429
Joined: Tue Sep 06, 2011 12:57 am
Location: Anchorage, AK

Re: Great Combo - BFM DSR280s & SAC

#22 Post by SirNickity »

Right now, the Intel platform allows any core in a multicore system instant access to RAM.
I want to know more. Are you saying that you are likely to run into access violations because of parallel threads reading/writing the same area of memory? Aren't locking mechanisms supposed to avoid that? (BTW: Interest, not challenge.)
Windows is quite stable... until you start gomming it up with all manner of crum.
Generally, this is true. Reliability is almost always a matter of driver stability and other things you have going on. That said, Microsoft has an infuriating habit of dumping way too much code into a single binary, resulting in tons of complexity. Code that tangled can never be bug-free. This is why (server perspective) I avoid having Windows exposed to the Internet at all costs. Also, each release of Windows has more of the aforementioned gom and crum built-in. I started running a Hackintosh because I was having a hard time getting XP stable on my last build, and the extra fluff in Win 7 is filled with crap I don't want on a real-time performance PC. (Sidebar: Has anyone else used the dev preview of Windows 8 yet? What a nightmare. Totally unsuited for this type of work.)

I would love to see a purpose-built, pro audio-targeted Linux distribution. Linux, as it stands now, is tough because it's an unknown quantity. Anyone can DIY their own flavor and there's zero consistency. But if a vendor released an easy-to-use, black-box release that had just the necessary bits and pieces, with a very sparse window manager, that would be absolutely unbeatable.
I can run 48 channels and 10 monitor mixes with multiple verbs, comps, eqs, gates, VST plugs, etc.... and on my E5200 system (not the fastest in the world) I top out at about 30-40%.
I always hear this, but I've had a hard time matching it. For example, I set up a Waves Multirack box running about 18 channels with compression, EQ, and a couple reverbs for sends. At ASIO buffer settings that didn't cause audible delay, I started getting dropouts and static in the audio. My CPU usage was at about 10-12%, but I think I ran out of memory bandwidth trying to shuttle 32 samples in, through the DSP, and back out in few enough ms. (This was on a Mobile Core 2 Quad 2.4GHz Mini-ITX build with a MOTU PCI-424 and 24i/o.) Also, the vocalist hated the phasing caused by the remaining delay in the monitor feed combined with his voice. I don't think any digital system can overcome that, at least until the delay is within a half-dozen samples.

Also, multitrack mixes of any size in Adobe Audition 3 would start bogging down, but I'm not convinced that Audition is terribly well written.

So, how does SAC fair with latency? I'm curious.

User avatar
BrentEvans
Posts: 3041
Joined: Thu Oct 09, 2008 10:38 am
Location: Salisbury, NC

Re: Great Combo - BFM DSR280s & SAC

#23 Post by BrentEvans »

SirNickity wrote:I want to know more. Are you saying that you are likely to run into access violations because of parallel threads reading/writing the same area of memory? Aren't locking mechanisms supposed to avoid that? (BTW: Interest, not challenge.)
As I understand it, that's it in a nutshell. Remember that in order to gain the low latency realtime operation, SAC's engine is coded in assembly, so presumably it operates under the radar of the locking mechanisms... or maybe it's the locking mechanisms themselves that cause the glitch. That's a question for the developer (just ask it nicely).

Generally, this is true. Reliability is almost always a matter of driver stability and other things you have going on. That said, Microsoft has an infuriating habit of dumping way too much code into a single binary, resulting in tons of complexity. Code that tangled can never be bug-free. This is why (server perspective) I avoid having Windows exposed to the Internet at all costs. Also, each release of Windows has more of the aforementioned gom and crum built-in. I started running a Hackintosh because I was having a hard time getting XP stable on my last build, and the extra fluff in Win 7 is filled with crap I don't want on a real-time performance PC. (Sidebar: Has anyone else used the dev preview of Windows 8 yet? What a nightmare. Totally unsuited for this type of work.)
Well... MS has kind of been like Star Trek Movies since win 98.. every other one is a good one. 98, good. Me, bad. XP, Good. Vista, Bad. 7, Good... I expect 8 will be to 7 as Vista was to XP.
I would love to see a purpose-built, pro audio-targeted Linux distribution. Linux, as it stands now, is tough because it's an unknown quantity. Anyone can DIY their own flavor and there's zero consistency. But if a vendor released an easy-to-use, black-box release that had just the necessary bits and pieces, with a very sparse window manager, that would be absolutely unbeatable.

The biggest problem with Linux for audio is poor driver support from pro level audio cards. Some have drivers, but they're not as good as the Windows and Mac drivers, because there really isn't much pro-level AV software on Linux. Sadly, Nix in general lends itself toward engineers and enthusiasts more than to the market... because there is no "standard" per se.
I always hear this, but I've had a hard time matching it. For example, I set up a Waves Multirack box running about 18 channels with compression, EQ, and a couple reverbs for sends. At ASIO buffer settings that didn't cause audible delay, I started getting dropouts and static in the audio. My CPU usage was at about 10-12%, but I think I ran out of memory bandwidth trying to shuttle 32 samples in, through the DSP, and back out in few enough ms. (This was on a Mobile Core 2 Quad 2.4GHz Mini-ITX build with a MOTU PCI-424 and 24i/o.) Also, the vocalist hated the phasing caused by the remaining delay in the monitor feed combined with his voice. I don't think any digital system can overcome that, at least until the delay is within a half-dozen samples.
Waves is a poor example of using plugs in live audio. Most of their plugs have latency, and are very processor and ram hungry. SAC doesn't support any plug which reports more than 0 samples of latency. Sometimes a plug will have latency in its processing and still report 0 samples... they typically do this by sending no data for the first few cycles. Those work and sometimes the latency is audible, but most plugs don't work that way. For instance, out of my Waves pack, there are only about 4 or 5 plugs that SAC will use, aside from the analyzers. The reverbs work OK, and I think there is one compressor that works (forget which one). They all hog CPU time too much so I use none of them, and there are others that sound as good or better (and are far less expensive, or free).
Also, multitrack mixes of any size in Adobe Audition 3 would start bogging down, but I'm not convinced that Audition is terribly well written.
That's Audition, not the mixes. Audition is an audio editor with a DAW shoehorned in, it's not well tuned, and it's Adobe - not exactly a DAW company.
So, how does SAC fair with latency? I'm curious.
Technically, SAC itself (as in the software) has very minimal or no latency. The system latency is entirely dependent upon your PC and audio hardware. Ultimately, SAC will operate at the minimum latency your hardware can tolerate. An average SAC system is 5-7ms, but could conceivably be as low as 2-3ms with enough horsepower and the right interface. This may sound like a lot, but it is well in line with the average digital mixer plus digital snake plus digital FOH processing that is used on live concerts. It only comes into play when mixing IEMs, and some of the more picky will hear phasing in their IEM mix... the solution is to either add a touch of vocal reverb to the iem mix, or increase the latency slightly with a delay plug to more closely emulate what they'd hear out of a wedge. Which one works better depends on the person.

IOW, SAC's latency is a non-issue for most users.
Last edited by BrentEvans on Sun Oct 02, 2011 10:48 am, edited 1 time in total.
99% of the time, things that aren't already being done aren't being done because they don't work. The other 1% is split evenly between fools and geniuses.

bassmonster
Posts: 1384
Joined: Wed Jul 13, 2011 9:23 am

Re: Great Combo - BFM DSR280s & SAC

#24 Post by bassmonster »

SirNickity wrote:Reliability is almost always a matter of driver stability and other things you have going on. That said, Microsoft has an infuriating habit of dumping way too much code into a single binary, resulting in tons of complexity. Code that tangled can never be bug-free. This is why (server perspective) I avoid having Windows exposed to the Internet at all costs. Also, each release of Windows has more of the aforementioned gom and crum built-in. I started running a Hackintosh because I was having a hard time getting XP stable on my last build, and the extra fluff in Win 7 is filled with crap I don't want on a real-time performance PC. (Sidebar: Has anyone else used the dev preview of Windows 8 yet? What a nightmare. Totally unsuited for this type of work.)
Ever had the problem in Windows 7 with the sidebar.exe application? There is a memory leak in mine, I once caught it hogging a WHOLE GB of RAM!!! I think it's a problem with the CPU meter gadget I was running, I just don't use it any more. +1 on the Linux pro audio distribution idea.

User avatar
Tom Smit
Posts: 7566
Joined: Sat Jan 05, 2008 1:24 pm
Location: Sarnia, Ont. Canada

Re: Great Combo - BFM DSR280s & SAC

#25 Post by Tom Smit »

Hey Brent, I just want to thank you for your constant info, and the plugging, of SAC. I want it, but I am way, way, way off in the sidelines. So any info you give helps me process (even though I could spend hours and hours reading all the forum at SAC/SAW)


Image
TomS

SirNickity
Posts: 429
Joined: Tue Sep 06, 2011 12:57 am
Location: Anchorage, AK

Re: Great Combo - BFM DSR280s & SAC

#26 Post by SirNickity »

SAC doesn't support any plug which reports more than 0 samples of latency.
I didn't think that was possible for many types of plugins. EQ is temporal, for instance. How do you do that without at least a few samples latency? Dynamics processing doesn't necessarily have to work on a range of time, but if you use any prediction it would. Though look-ahead is probably off the table for real-time processing anyway.
Hey Brent, I just want to thank you for your constant info, and the plugging, of SAC.
Seconded. :-) Thanks for the discussion.
Ever had the problem in Windows 7 with the sidebar.exe application?
YES. I reboot like once every couple months, but yeah, I regularly have to kill sidebar.exe because the displays wig out and it pegs one of my cores at 100%.

BoostFab
Posts: 458
Joined: Fri Sep 17, 2010 2:51 pm
Location: DFW, TX

Re: Great Combo - BFM DSR280s & SAC

#27 Post by BoostFab »

SAC is way a head of its game in the digital mixing industry, even 10 years ago. Now we see the big names are moving to PC controlled and wireless remote platform just like SAC. I believe eventually Ryder will accept SAC. I'm moving from a 24ch. SoundCraft mixer to SAC, it allows me to get rid of a rack full of hardware to a smaller footprint with 100x more capability.

User avatar
BrentEvans
Posts: 3041
Joined: Thu Oct 09, 2008 10:38 am
Location: Salisbury, NC

Re: Great Combo - BFM DSR280s & SAC

#28 Post by BrentEvans »

Tom Smit wrote:Hey Brent, I just want to thank you for your constant info, and the plugging, of SAC. I want it, but I am way, way, way off in the sidelines. So any info you give helps me process (even though I could spend hours and hours reading all the forum at SAC/SAW)
No problem... jump in... the water's fine!
SirNickity wrote: I didn't think that was possible for many types of plugins. EQ is temporal, for instance. How do you do that without at least a few samples latency? Dynamics processing doesn't necessarily have to work on a range of time, but if you use any prediction it would. Though look-ahead is probably off the table for real-time processing anyway.
EQ is temporal only in that it affects phase, so the affects are only within 360 degrees. This can be handled within a sample, so doesn't require latency. Lookahead processing does require latency, but typically isn't necessary.
99% of the time, things that aren't already being done aren't being done because they don't work. The other 1% is split evenly between fools and geniuses.

Gregory East
Posts: 3495
Joined: Sat May 13, 2006 9:56 pm

Re: Great Combo - BFM DSR280s & SAC

#29 Post by Gregory East »

I had to google "lookahead" but that's almost the very definition of "latency" isn't it? An inbuilt constant delay from signal generation to reproduction?
BAT10, Bad Auto Tuba. Reverse folded TAT to fit JBL 1014D, 350W driver, voltage limit unknown.

Post Reply