HDX And DSP Cache

[Emcha_audio]

That's actually not new. I think it was even mentioned in the original presentation of the HDX card at aes.

[Bushpig]

Manny,

I think it was even mentioned in the original presentation of the HDX card at aes.

You may be right. I didn't spot that at the time. Still I know I'm not the only one who HATES working with playlists. Not least because on the rare occasions I do use them, I've several times had the bug where grouped tracks that had their playlists created simultaneously, refuse to switch as a group afterwards. Drives me crazy and on more than one occasion has got me in a right pickle when I didn't notice they didn't switch together.

Then there's the issue of doing full song edits with the "All" group enabled, but the underlying playlists don't follow. I know this drives some of the film guys crazy.

Plus, I put one of our fellow users here onto the "voice stealing Instacomp®" method recently (can't remember who it was) after he'd been struggling with comp'ing using playlists, and he said the first couple of his regular clients that came in, loved it. It IS the most efficient way to work IMHO. Tidier, easier to keep track of etc..

Oh well, another good reason to stick with the ol' TDM system and ride it into the sunset. No money for Avid from me. I swear these guys have never worked on a "real" session. They all claim to be users of the software, but for frick's sake, they've got day jobs, that makes them amateurs and certainly doesn't subject them to the same client facing pressures that we all face when the software doesn't behave like it should.

Still, I do love the software, just wish I wasn't being crippled by EOL (years of 32 bit memory crap and shxxty RTAS VI performance) on the older hardware.

Cheers all.

Steve Bush
www.music180.com/pros/5887

MacPro 4.1 (Nehalem) 2.26 8 Core, OSX 10.6.8, PT10HD, 14 Gig RAM, PCIe HD3, 192, Sync I/O, Midi I/O.

[Emcha_audio]

Originally Posted by Bushpig

Manny,



You may be right. I didn't spot that at the time. Still I know I'm not the only one who HATES working with playlists. Not least because on the rare occasions I do use them, I've several times had the bug where grouped tracks that had their playlists created simultaneously, refuse to switch as a group afterwards. Drives me crazy and on more than one occasion has got me in a right pickle when I didn't notice they didn't switch together.

Then there's the issue of doing full song edits with the "All" group enabled, but the underlying playlists don't follow. I know this drives some of the film guys crazy.

Plus, I put one of our fellow users here onto the "voice stealing Instacomp®" method recently (can't remember who it was) after he'd been struggling with comp'ing using playlists, and he said the first couple of his regular clients that came in, loved it. It IS the most efficient way to work IMHO. Tidier, easier to keep track of etc..

Oh well, another good reason to stick with the ol' TDM system and ride it into the sunset. No money for Avid from me. I swear these guys have never worked on a "real" session. They all claim to be users of the software, but for frick's sake, they've got day jobs, that makes them amateurs and certainly doesn't subject them to the same client facing pressures that we all face when the software doesn't behave like it should.

Still, I do love the software, just wish I wasn't being crippled by EOL (years of 32 bit memory crap and shxxty RTAS VI performance) on the older hardware.

Cheers all.

Steve Bush
www.music180.com/pros/5887

MacPro 4.1 (Nehalem) 2.26 8 Core, OSX 10.6.8, PT10HD, 14 Gig RAM, PCIe HD3, 192, Sync I/O, Midi I/O.

Yeah I get what you mean. What I remember from the presentation is that they said that it was a dynamic voice allocation system. I'm not sure if it pertains to it or not.

[Drew Mazurek]

Originally Posted by Bushpig

Not least because on the rare occasions I do use them, I've several times had the bug where grouped tracks that had their playlists created simultaneously, refuse to switch as a group afterwards. Drives me crazy and on more than one occasion has got me in a right pickle when I didn't notice they didn't switch together.
MacPro 4.1 (Nehalem) 2.26 8 Core, OSX 10.6.8, PT10HD, 14 Gig RAM, PCIe HD3, 192, Sync I/O, Midi I/O.

FWIW, I've NEVER seen this if the Group was created BEFORE you've recorded anything and I've seen it often when you try to make the group after having already recorded something. For me, I see a major correlation.

[PaulV]

Regarding DSP Cache, its main function was work around HD plug-in loading being slow. Plug-in loading in HDX is orders of magnitude faster, so we didn't feel it was worth the added complexity.

Regarding workflows involving the old voice sharing techniques, we're actively working on solutions in this area. We'll keep you updated.

[Touchwood Studios]

Still it would be great to have a choice to turn it off.
I hate it when I loose features ( Like Apple does to us ).....
The voice thing does no bother me at all 512 voices on 2 cards is plenty :)

[Bushpig]

Morning all.

Drew,

I've NEVER seen this if the Group was created BEFORE you've recorded anything

Aahh, that's interesting. Thinking back over the times I've fallen victim to this behaviour, I don't think I've ever created playlists on my original grouped recording tracks. Normally I'm grouping some kind of alternative comp tracks and wanting to dupe the playlist on those before I try an alternate edit etc, but then I fail to notice when I switch back to the original playlist, that it sometimes hasn't switched all the tracks in unison. I've messed up some stuff badly because of this and it's made me highly suspicious of playlists in general. Might be an over reaction, but when we're at the coal face of client facing pressure, it simply can't be relied upon without being highly attentive, and I just don't need that sort of extra concentration overhead if I can avoid it. Hence the "voice stealing Instacomp" method I'm so fond of.

PaulV,

so we didn't feel it was worth the added complexity.

Sounds like a thoroughly sensible approach to me. The less complex the better I'd say.

Regarding workflows involving the old voice sharing techniques, we're actively working on solutions in this area. We'll keep you updated.

Thanks for that info. Any major technical reason you can share with us as to why voice sharing was abandoned with HDX??

Touchwood,

The voice thing does no bother me at all 512 voices on 2 cards is plenty

Yes agreed, this is a great thing, but that's not the issue that will be problematic for those of us who use the "voice stealing" workflow. It may not be something that you've tried, but it's a beautiful thing for some of us, and will be sorely missed if it has been killed off completely.

Cheers all.

Steve Bush
www.music180.com/pros/5887

2 x Systems: MacPro 4.1 (Nehalem) 2.26 8 Core, OSX 10.6.8, PT10HD, 14 Gig RAM, PCIe HD3, 192's, Sync I/O, Midi I/O.

[WernerF]

Hey Steve, I know that this suggestion isn't an answer to totally replace your voice stealing workflow, but with so many extra voices being available in HDX, couldn't you just record takes to separate tracks and then just comp from them? Of course there're some extra muting moves involved but at least you wouldn't have to trust playlists.

[PaulV]

Any major technical reason you can share with us as to why voice sharing was abandoned with HDX?

Yes, since you can have a different mix on tracks sharing a voice, sharing voices requires timed switching of the mixer during playback so the shared audio stream can be rerouted sample-accurately to another mixer channel whenever a track 'drops-through'. HDX doesn't do this currently - it's a complex piece of work to add for a feature that was originally created to get around not having enough voices / tracks.

That said, hopefully we can add the drop-through track comping features using a different approach.

Voice sharing was provided in Pro Tools 1.0 when you only had 4 voices and squeezing every last bit of bandwidth out of the interface was really important. The voicing was basically a hardware detail that was exposed to the user for working around its limitations. But over the years, folks have come up with creative ways to use this, such as track comping and a handful of post-production techniques.

With HDX, the voice limitations are more or less gone (well, there are still a few post folks complaining about 768 tracks not being enough; but we're working on it!), but HD users still like the things that could be done with voice sharing, so we'll try to open up those workflows again.

We'll keep you posted.

[Drew Mazurek]

Can I just say that this sort of exchange between users and Avid engineering is awesome!! Would be great to have more of it! Plus I am learning a lot!