Anyone used either of these?
Sound Sculpture Switchblade
CM Labs SixtyFour
I'm seriously considering 'em, leaning toward the Switchblade GL at this point. Comments would be much appreciated.
Thanks!
Johann
MIDI-controllable patchbays
-
blazesboylan
- Posts: 777
- Joined: Sat May 25, 2002 4:00 pm
- Location: The Great White North
- Contact:
-
blazesboylan
- Posts: 777
- Joined: Sat May 25, 2002 4:00 pm
- Location: The Great White North
- Contact:
They each go for $1800 USD.
They each have their advantages, but I haven't been able to find anyone who's used them (except for the Harmoron Central crowd).
Of course you can do all this stuff in SFP, but I'm fighting tooth and nail to get every last drop of DSP out of my system so... These are tempting.
If anyone has experience please do share! Otherwise I'll post whatever (if anything) I find out...
Cheers,
Johann
They each have their advantages, but I haven't been able to find anyone who's used them (except for the Harmoron Central crowd).
Of course you can do all this stuff in SFP, but I'm fighting tooth and nail to get every last drop of DSP out of my system so... These are tempting.
If anyone has experience please do share! Otherwise I'll post whatever (if anything) I find out...
Cheers,
Johann
Are your signals already in the analog realm? I'm curious because you say that SFP can do this, but I'm wondering exactly what it is you're trying to do? If I recall correctly you had a massive SFP project set up for routing & mixing.
<font size=-1>[ This Message was edited by: valis on 2004-07-29 21:52 ]</font>
<font size=-1>[ This Message was edited by: valis on 2004-07-29 21:52 ]</font>
-
blazesboylan
- Posts: 777
- Joined: Sat May 25, 2002 4:00 pm
- Location: The Great White North
- Contact:
Hi Valis.
Routing buses and L/R master to external compressors.
Possibly also useful for aux buses (send external FX unit 1 output to unit 2, etc).
A bit of background for anyone who's following along
: the absurdly big project I'm working on is a 32x8x2 +6aux mixer with most of the dynamics and EQ etc controllable by MIDI.
I currently have 2 SFP mixers providing "insert matrix" functionality:
I had to remove a couple of internal compressors + effects. 
After my initial excitement over the external MIDI-controllable patchbays earlier today, I'm beginning to think it would be wiser to buy a few more external compressors and effects units. I could just use SFP for routing and replace all internal dynamics / FX with external ones. $1800 is a lot of external processing...
Cheers,
Johann
P.S. I do plan on uploading the absurd project here once I've stabilized it... Smaller versions, too.
P.P.S. I'm using the Tcl scripting language to send MIDI controls from a crappy old Linux box. I'll upload the Tcl modules here too. If anyone reading this is interested in automating their CreamWare rig remotely, start reading up on Tcl / Tk.
Routing buses and L/R master to external compressors.
Possibly also useful for aux buses (send external FX unit 1 output to unit 2, etc).
A bit of background for anyone who's following along
: the absurdly big project I'm working on is a 32x8x2 +6aux mixer with most of the dynamics and EQ etc controllable by MIDI.I currently have 2 SFP mixers providing "insert matrix" functionality:
- 1 "patchbay" matrix for the L/R master bus and buses 3-8 to / from internal and external compressors. (I leave buses 1 and 2 as mono with fixed comps, un-re-routable.)
- 1 matrix for feeding back the external effects into each other. (Reason: only 1 aux send allowed per channel strip. Therefore "aux - to - aux" sends are somewhat necessary.)
I had to remove a couple of internal compressors + effects. After my initial excitement over the external MIDI-controllable patchbays earlier today, I'm beginning to think it would be wiser to buy a few more external compressors and effects units. I could just use SFP for routing and replace all internal dynamics / FX with external ones. $1800 is a lot of external processing...
Cheers,
Johann
P.S. I do plan on uploading the absurd project here once I've stabilized it... Smaller versions, too.
P.P.S. I'm using the Tcl scripting language to send MIDI controls from a crappy old Linux box. I'll upload the Tcl modules here too. If anyone reading this is interested in automating their CreamWare rig remotely, start reading up on Tcl / Tk.
I guess what I'm really wondering is: is everything you're mixing digital? My impression was that was probably the case (at least that the stuff gets tracked into the DAW) and I would recommend against several ad/da conversions.
Another suggestion is to give smaller SFP mixers a go and try to figure out how to work that into your overall setup. I use to use nothing but the largest stereo mixer (the one that came with pro pack) and was always running out of dsp on a 6 dsp card. Eventually I just started using micromixer and chaining off the inputs for any complicated routing, if you require direct outs as well as summing in the mixer I think the smallest stm does support that? The important thing is to have something that dynamically allocates resources (micromixer and dynamixer from what I'm aware) and do all your processing with external modules. This gives you total control over the phase errors that can occur (and they will occur with inserts even with the compensation enabled on the mixer the effect is inserted in) and it also seems to drastically reduce dsp usage.
A comment on mixing strategies. First I will mention that SFP is no longer my main ASIO host so things have changed slightly, but let's ignore that until the end and assume SFP is getting your software's output directly into its asio drivers.
Whether mixing in Logic, Sx, SFP or my hardware mixer I can definately hear audible differences. However rather than viewing one as superior or others as inferior, I just recognize that I'm going to get a sonic 'print' depending on where things are summed. I think of my mixing in 'stages'. Instead demanding an individual output for every track and trying to sum in just 1 stereo master bus, I think of things in terms of 'signal chains' and mix accordingly. Some things are ok summed in software, especially if you want them to act as the same layer/part in the mix. Often I will even output things from my software on 1 channel if they're distinct parts (have thier own individual places in time) but they're sonically similar somehow (all pads etc). Things that I definately want distinctly separated or processed elsewhere (sfp or external) I output individually of course.
But much as a digital synth that allows you to combine 4 base 'tones' (like much of the korg & roland digital synths) into a single 'part' I see no reason not to work this way in mixing. When things come into SFP I will often have a 'chain' that has many parts combined, so any compression or other processing here will affect things as a unit. Again if I want to treat a sound separately I may decide that chain (my basses for example) should be more separated as it enters SFP and then treat the individual parts I want there and sum them before adding them into the total mix. In fact my normal SFP project uses several 'submixers' and a larger 'main' mixer. I will often even have dedicated mixers for reverb etc that take inputs chained off elsewhere and feed them into that mixer to control relative 'send' levels into the mixer, then just route the output of that mixer through the reverb (set to 100% wet, dry off) and then feed the reverb into the appropriate sub/main mixer and have the channel volume as my 'return'.
Now when it gets to my outboard mixer, it has 2 uses in life really: either to record my external synths (via a bus output typically) or to add a bit of 'grit' to a signal 'chain' when doing the final mix. So you could say that my analog mixer is used to sum the SFP based 'submixers'. Obviously before I had the hardware mixer I simply summed into another SFP mixer, but this mixer has inputs that overdrive nicely and allow a bit more 'grit' than I was able to get out of the digital realm so I mix there now. Again if I want, say, my kick & snare 'hot' and gritty, but the rest of the 'percussion' to just sit nicely in the mix, I send them to separate inputs on my mixer from SFP's outputs and treat them separately there. I may even bus all my drums or basses to a final analog bus and compress there, but I would rarely redigitize them and run them back through unless I wanted to take advantage of what 'resampling' would do for you in in the pre-software days (ie, on an emu resampling several layered parts to treat them again as 1 thing).
If you're recording everything into software and playing it back from there (classic tracking), or using a more modern combination of audio tracks, vsti's and creamware synthesis & fx you can think of it this way:
The first stage in mixing is what is in your software, audio tracks (recorded) or software instruments (live/frozen/bounced) combine into. Rather than needing 64 outputs by judiciously choosing what is summed in software I usually need about 10-18 ASIO outs and 3-4 'submixers' or main mix stereo 'busses' (of course I write drum 'n bass without lead vocals typically).
The second stage of mixing is what happens in SFP, and for most digital users this is the final stage. But again only those things that are most important or that you want to specifically treat in SFP need to be separate here. Often for example my kick/snare/hat will come out of software as separate things (compressed for snappy attack in software) and summed in a micromixer (perhaps with a nice bit of sfp verb added & eq'd into the same mixer as a 'send return') and then they get compressed as a unit (bus compressed) before being sent to the master SFP mixer (or now my outboard analog mixer).
The third stage is either going to besumming your final bus 'chains' and 'returns' into a master SFP mixer, or in my case my outboard mixer again). Working in this way means I only need 16 channels (tho I have 8 busses as well) on my main analog mixer, and I typically only need 3-4 micromixers in sfp.
Now I mentioned earlier that I use another ASIO driver these days, I have an RME hammerfall/multiface combination and have 8 i/o available there to get hardware in and my mixing out to my main mixer. I also have the individual i/o on the Creamware cards, but my dsp cards are now in an old P3 machine. This means I can route some tracks to a Logic output that goes to (for example) track 5 & 6 on my analog mixer (directly out of the multiface).
Yet something else I may want to pass through a nice SFP effect or do some bus 'compression' in there with Vinco rather than in software so I will use a separate Logic output routed through hammerfall's totalmix to the adat ins on my Pulsar2. That will then be treated and sent back out the Pulsar2's adat outs and then to the SAME output from my multiface into track 5 & 6 on my analog mixer. So effectively I'm summing inside RME's totalmix and sending them as 1 analog 'track' yet still treating things as distinct signal 'chains'.
Another example would be to basses. I typically use a synthsized sin wave or heavily filtered saw for my 'sub' and layer at least 3-4 other bass 'sounds' on top of that. So I may have a simple vsti generating the sub (discodsp's discovery has good antialiasing for this), and then have another vsti or two added in and send them all (the sub and the layered bassed) to SFP. Then I might choose a sound on my Virus (outboard) and another on my basstation and route them INTO the hammerfall then the SFP via adat. Sum them in SFP to compress & treat them with something, then route them back out the adat ports through my hammerfall & out an analog out to just 1-2 tracks on my analog mixer.
Roughly broken downI might have 10-20 audio 'tracks', another 10-15 software instrument outputs (mostly exs24 on drums & percussion in software), 2-3 sfp synths (minimax/pro one etc) and 3-4 outboard synths, all mixed and processed at various points in my signal chain as appropriate.
Another suggestion is to give smaller SFP mixers a go and try to figure out how to work that into your overall setup. I use to use nothing but the largest stereo mixer (the one that came with pro pack) and was always running out of dsp on a 6 dsp card. Eventually I just started using micromixer and chaining off the inputs for any complicated routing, if you require direct outs as well as summing in the mixer I think the smallest stm does support that? The important thing is to have something that dynamically allocates resources (micromixer and dynamixer from what I'm aware) and do all your processing with external modules. This gives you total control over the phase errors that can occur (and they will occur with inserts even with the compensation enabled on the mixer the effect is inserted in) and it also seems to drastically reduce dsp usage.
A comment on mixing strategies. First I will mention that SFP is no longer my main ASIO host so things have changed slightly, but let's ignore that until the end and assume SFP is getting your software's output directly into its asio drivers.
Whether mixing in Logic, Sx, SFP or my hardware mixer I can definately hear audible differences. However rather than viewing one as superior or others as inferior, I just recognize that I'm going to get a sonic 'print' depending on where things are summed. I think of my mixing in 'stages'. Instead demanding an individual output for every track and trying to sum in just 1 stereo master bus, I think of things in terms of 'signal chains' and mix accordingly. Some things are ok summed in software, especially if you want them to act as the same layer/part in the mix. Often I will even output things from my software on 1 channel if they're distinct parts (have thier own individual places in time) but they're sonically similar somehow (all pads etc). Things that I definately want distinctly separated or processed elsewhere (sfp or external) I output individually of course.
But much as a digital synth that allows you to combine 4 base 'tones' (like much of the korg & roland digital synths) into a single 'part' I see no reason not to work this way in mixing. When things come into SFP I will often have a 'chain' that has many parts combined, so any compression or other processing here will affect things as a unit. Again if I want to treat a sound separately I may decide that chain (my basses for example) should be more separated as it enters SFP and then treat the individual parts I want there and sum them before adding them into the total mix. In fact my normal SFP project uses several 'submixers' and a larger 'main' mixer. I will often even have dedicated mixers for reverb etc that take inputs chained off elsewhere and feed them into that mixer to control relative 'send' levels into the mixer, then just route the output of that mixer through the reverb (set to 100% wet, dry off) and then feed the reverb into the appropriate sub/main mixer and have the channel volume as my 'return'.
Now when it gets to my outboard mixer, it has 2 uses in life really: either to record my external synths (via a bus output typically) or to add a bit of 'grit' to a signal 'chain' when doing the final mix. So you could say that my analog mixer is used to sum the SFP based 'submixers'. Obviously before I had the hardware mixer I simply summed into another SFP mixer, but this mixer has inputs that overdrive nicely and allow a bit more 'grit' than I was able to get out of the digital realm so I mix there now. Again if I want, say, my kick & snare 'hot' and gritty, but the rest of the 'percussion' to just sit nicely in the mix, I send them to separate inputs on my mixer from SFP's outputs and treat them separately there. I may even bus all my drums or basses to a final analog bus and compress there, but I would rarely redigitize them and run them back through unless I wanted to take advantage of what 'resampling' would do for you in in the pre-software days (ie, on an emu resampling several layered parts to treat them again as 1 thing).
If you're recording everything into software and playing it back from there (classic tracking), or using a more modern combination of audio tracks, vsti's and creamware synthesis & fx you can think of it this way:
The first stage in mixing is what is in your software, audio tracks (recorded) or software instruments (live/frozen/bounced) combine into. Rather than needing 64 outputs by judiciously choosing what is summed in software I usually need about 10-18 ASIO outs and 3-4 'submixers' or main mix stereo 'busses' (of course I write drum 'n bass without lead vocals typically).
The second stage of mixing is what happens in SFP, and for most digital users this is the final stage. But again only those things that are most important or that you want to specifically treat in SFP need to be separate here. Often for example my kick/snare/hat will come out of software as separate things (compressed for snappy attack in software) and summed in a micromixer (perhaps with a nice bit of sfp verb added & eq'd into the same mixer as a 'send return') and then they get compressed as a unit (bus compressed) before being sent to the master SFP mixer (or now my outboard analog mixer).
The third stage is either going to besumming your final bus 'chains' and 'returns' into a master SFP mixer, or in my case my outboard mixer again). Working in this way means I only need 16 channels (tho I have 8 busses as well) on my main analog mixer, and I typically only need 3-4 micromixers in sfp.
Now I mentioned earlier that I use another ASIO driver these days, I have an RME hammerfall/multiface combination and have 8 i/o available there to get hardware in and my mixing out to my main mixer. I also have the individual i/o on the Creamware cards, but my dsp cards are now in an old P3 machine. This means I can route some tracks to a Logic output that goes to (for example) track 5 & 6 on my analog mixer (directly out of the multiface).
Yet something else I may want to pass through a nice SFP effect or do some bus 'compression' in there with Vinco rather than in software so I will use a separate Logic output routed through hammerfall's totalmix to the adat ins on my Pulsar2. That will then be treated and sent back out the Pulsar2's adat outs and then to the SAME output from my multiface into track 5 & 6 on my analog mixer. So effectively I'm summing inside RME's totalmix and sending them as 1 analog 'track' yet still treating things as distinct signal 'chains'.
Another example would be to basses. I typically use a synthsized sin wave or heavily filtered saw for my 'sub' and layer at least 3-4 other bass 'sounds' on top of that. So I may have a simple vsti generating the sub (discodsp's discovery has good antialiasing for this), and then have another vsti or two added in and send them all (the sub and the layered bassed) to SFP. Then I might choose a sound on my Virus (outboard) and another on my basstation and route them INTO the hammerfall then the SFP via adat. Sum them in SFP to compress & treat them with something, then route them back out the adat ports through my hammerfall & out an analog out to just 1-2 tracks on my analog mixer.
Roughly broken downI might have 10-20 audio 'tracks', another 10-15 software instrument outputs (mostly exs24 on drums & percussion in software), 2-3 sfp synths (minimax/pro one etc) and 3-4 outboard synths, all mixed and processed at various points in my signal chain as appropriate.
-
blazesboylan
- Posts: 777
- Joined: Sat May 25, 2002 4:00 pm
- Location: The Great White North
- Contact:
Valis:
The signals are not all digital. The simplified version of the mixer looks like:
However this is overlooking the fact that there is external gear. The 8 buses and L/R mix bus all have compressors on them. Buses 1, 3, 4, L/R currently use external compression. Auxes 1, 2, 3 currently use outboard effects. I plan on adding outboard compression to buses 5/6 soon, and 2, 7, 8 eventually. Auxes 4-6 will also eventually be external effects units.
The ultimate goal is to have all 8 buses and L/R routed to outboard compressors via the virtual patchbay. And the Aux goal is to have all 6 auxes routed to external FX units, which get routed back into the feedback matrix so that, for example, a delay can be fed to a reverb, or a flanger to a tremolo, etc.
I have not started dealing with latency, but you are wise to advise that it will be an issue. If I were to, for example, send a kick drum out on channel 1 and route it to bus 3 through an external compressor back into the mix as well as to aux 1, through an effects unit into the mix, there would be some serious lag time. I have not dealt with this yet but I will have to.
I've tried many configs and I am thoroughly convinced that the 48 / X mixers are the most DSP-efficient creations imaginable. They do have to be tweaked heavily. But they are infinitely more efficient than any other mixer or other device I have tried using. I've found that the connections between devices cost much more than the routing within a mixer.
The key for me is to avoid touching SFP and sequencer GUIs altogether. Remote control is the be-all and end-all of this project. Even remote control of external devices. I'm currently wrestling with sending bank + program messages to the FX units. ALSA is not my friend at the moment.
Neverthelesss I understand that there are many opinions on which sequencers "sound best". I have a sneaking suspicion that it's all horsesh** but I am nevertheless happy mixing in SFP. I was, for a long time, afraid of moving from analog to digital. But it sounds good and *clean* to me, otherwise I wouldn't have moved. Dirt is not something I look for in a mixer! Compressors and FX add plenty of that.
In terms of summing I don't hear differences between sequencers. But EQ and compression and so on sound like crap in every sequencer I've ever used. SFP EQ and compression sound transparent to me, and that has been my criterion in deciding that the time is right to automate.
So my outboard mixer (a clean sounding but overly hated Mackie 8 bus) has become an oversized mic pre / headphone amp.
In mixing rock'n'roll I've never used more than 30 tracks. Therefore I figure a 32x8x2 mixer with 6 auxes is a reasonable goal. I originally hoped for 48x16x2, but that has turned out to be quite impossible without either: 1) a substantial increase in processing power; or 2) a brain that has an order of magnitude more imagination than mine in setting up the SFP project.
FWIW your idea of the "SFP effects device" is interesting to me. A cheap computer with a Pulsar II / Scope Project card is much cheaper than a really high-end effects device. And I suspect that if the presets were created carefully, it would sound better, too.
The only problem is that there are no pre-made project files that can be used as multi-effects. Somebody please correct this. Personally I would like to see a project file that includes maybe 2 multiFX modules with programs set up to provide lush reverb / chorus / EQ / etc combos. Like any multiFX device made by Lexicon, Eventide, Yamaha, now Kurzweil, and so on.
Valis I don't know whether you'd be interested in trying out this automated stuff once it's ready to go. But you seem to be the prime candidate -- your SFP setup isn't on your main machine. But what if you had 16-32 channels of EQ and dynamics that you could control from your RME machine. Would that be useful to you?
Cheers and sorry for taking so long to reply...!
Johann
<font size=-1>[ This Message was edited by: blazesboylan on 2004-08-04 02:34 ]</font>
The signals are not all digital. The simplified version of the mixer looks like:
However this is overlooking the fact that there is external gear. The 8 buses and L/R mix bus all have compressors on them. Buses 1, 3, 4, L/R currently use external compression. Auxes 1, 2, 3 currently use outboard effects. I plan on adding outboard compression to buses 5/6 soon, and 2, 7, 8 eventually. Auxes 4-6 will also eventually be external effects units.
The ultimate goal is to have all 8 buses and L/R routed to outboard compressors via the virtual patchbay. And the Aux goal is to have all 6 auxes routed to external FX units, which get routed back into the feedback matrix so that, for example, a delay can be fed to a reverb, or a flanger to a tremolo, etc.
I have not started dealing with latency, but you are wise to advise that it will be an issue. If I were to, for example, send a kick drum out on channel 1 and route it to bus 3 through an external compressor back into the mix as well as to aux 1, through an effects unit into the mix, there would be some serious lag time. I have not dealt with this yet but I will have to.
I've tried many configs and I am thoroughly convinced that the 48 / X mixers are the most DSP-efficient creations imaginable. They do have to be tweaked heavily. But they are infinitely more efficient than any other mixer or other device I have tried using. I've found that the connections between devices cost much more than the routing within a mixer.
The key for me is to avoid touching SFP and sequencer GUIs altogether. Remote control is the be-all and end-all of this project. Even remote control of external devices. I'm currently wrestling with sending bank + program messages to the FX units. ALSA is not my friend at the moment.
Neverthelesss I understand that there are many opinions on which sequencers "sound best". I have a sneaking suspicion that it's all horsesh** but I am nevertheless happy mixing in SFP. I was, for a long time, afraid of moving from analog to digital. But it sounds good and *clean* to me, otherwise I wouldn't have moved. Dirt is not something I look for in a mixer! Compressors and FX add plenty of that.
In terms of summing I don't hear differences between sequencers. But EQ and compression and so on sound like crap in every sequencer I've ever used. SFP EQ and compression sound transparent to me, and that has been my criterion in deciding that the time is right to automate.
So my outboard mixer (a clean sounding but overly hated Mackie 8 bus) has become an oversized mic pre / headphone amp.
In mixing rock'n'roll I've never used more than 30 tracks. Therefore I figure a 32x8x2 mixer with 6 auxes is a reasonable goal. I originally hoped for 48x16x2, but that has turned out to be quite impossible without either: 1) a substantial increase in processing power; or 2) a brain that has an order of magnitude more imagination than mine in setting up the SFP project.
FWIW your idea of the "SFP effects device" is interesting to me. A cheap computer with a Pulsar II / Scope Project card is much cheaper than a really high-end effects device. And I suspect that if the presets were created carefully, it would sound better, too.
The only problem is that there are no pre-made project files that can be used as multi-effects. Somebody please correct this. Personally I would like to see a project file that includes maybe 2 multiFX modules with programs set up to provide lush reverb / chorus / EQ / etc combos. Like any multiFX device made by Lexicon, Eventide, Yamaha, now Kurzweil, and so on.
Valis I don't know whether you'd be interested in trying out this automated stuff once it's ready to go. But you seem to be the prime candidate -- your SFP setup isn't on your main machine. But what if you had 16-32 channels of EQ and dynamics that you could control from your RME machine. Would that be useful to you?
Cheers and sorry for taking so long to reply...!
Johann
<font size=-1>[ This Message was edited by: blazesboylan on 2004-08-04 02:34 ]</font>
Well there are 2 things here.
1, The main thing that bothers me is that you're using EXTERNAL compression, which suggests to me that you're either feeding signals out & back in (or hopefully just mixing in the analog domain once you send out the 8 'busses'?).
2, in answer to your question about automation, I have moved over to using the pulsar more as a creative tool and less as a straight mixing tool. Meaning I route stuff to this machine from a Logic output, or I create signal chains entirely in SFP driven by midi to do what I want to do (heavily layered synthetic sounds etc). I'm always interested in trying new things with the gear I have though.
1, The main thing that bothers me is that you're using EXTERNAL compression, which suggests to me that you're either feeding signals out & back in (or hopefully just mixing in the analog domain once you send out the 8 'busses'?).
2, in answer to your question about automation, I have moved over to using the pulsar more as a creative tool and less as a straight mixing tool. Meaning I route stuff to this machine from a Logic output, or I create signal chains entirely in SFP driven by midi to do what I want to do (heavily layered synthetic sounds etc). I'm always interested in trying new things with the gear I have though.
-
hubird
-
blazesboylan
- Posts: 777
- Joined: Sat May 25, 2002 4:00 pm
- Location: The Great White North
- Contact:
The mixing is done in the digital domain. Yes phase / latency is going to have to be tackled!
I can tell you, though, that even with latency issues, the sound and flexibility of the hybrid approach (digital mixing / routing and analog signal processing) are fantastic. The outboard colouring really thickens things up in the mix.
Most of the outboard gear I'm using has only analogue I/O, but Hubird's S/PDIF "insert" trick works great for outboard effects that have that option!
Cheers,
Johann
I can tell you, though, that even with latency issues, the sound and flexibility of the hybrid approach (digital mixing / routing and analog signal processing) are fantastic. The outboard colouring really thickens things up in the mix.
Most of the outboard gear I'm using has only analogue I/O, but Hubird's S/PDIF "insert" trick works great for outboard effects that have that option!
Cheers,
Johann