The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
HD Native really Works!
Old 9th September 2011
  #31
Lives for gear
 
mykhal c's Avatar
 

RTAS plugs eat voices in TDM correct???...along with everythin' else. you do not lose voices in Native when doin' this. 192 voices in TDM is NOT the same 192 voices in Native. except for I/O count and TDM only plugs, a TDM system will be on the losin' end just about every time. low latency trackin'?? 32sample buffer should do it DSP on cards just cannot compete with today's systems.
Old 9th September 2011
  #32
Lives for gear
 
s.d.finley's Avatar
Quote:
Originally Posted by mykhal c View Post
RTAS plugs eat voices in TDM correct???...along with everythin' else. you do not lose voices in Native when doin' this. 192 voices in TDM is NOT the same 192 voices in Native. except for I/O count and TDM only plugs, a TDM system will be on the losin' end just about every time. low latency trackin'?? 32sample buffer should do it DSP on cards just cannot compete with today's systems.
RtAS on auxes and tdm then rtas plugs on audio tracks will eat voices in TDMville.
Old 9th September 2011
  #33
Lives for gear
 
barryjohns's Avatar
 

You can use tos same RTAS plugs n your TDM system. Best of both worlds. I always change TDM to RTAS after I'm done tracking on that track.
Old 9th September 2011
  #34
Some uses of RTAS use voices, yes, but it's blown out of proportion. I had/have an HD core test session with 80 audio tracks, a few auxes, and over 100 RTAS plugs (including plugs on the auxes) without exceeding the 96 voice count HD core limit. HD2 systems have the 192 voice limit, I'm talking only about an HD core.

I will email the session to anyone wanting it.

Native wins on voices, but voices are not the end all be all they're made out to be. I run 80-100 tracks with 100 RTAS routinely and voices are not the enemy - running out of delay compensation (only 4096 samples in PT) and CPU is the enemy.
Old 9th September 2011
  #35
Lives for gear
 
mykhal c's Avatar
 

but if everythin' in TDM makes one run outta/count voices, how can voices be blown outta proportion? the ONLY thing that uses voices in Native is track count. so while TDM users are addin' up track count, RTAS, TDM, delay comp, etc etc Native users are doin' all i just mentioned, not concerned, and only countin' tracks...all the while I/O is bein' handled on the card with the cpu free to handle all else.
Old 9th September 2011
  #36
Lives for gear
 
scott petito's Avatar
 

yup...

and i do understand the apples to apples comparison but i don't know of any way that could truly be done... too many variables...

my point for starting the thread was to confirm that under the right conditions ie a modern computer, enough ram and a native card you can get results that rival a tdm system does it rival an hd 8 ..no... but it appears in my case to at least equal or surpass my tdm system which is an HD 3.. and at the moment that includes tracking as well as mixing..

and I may be crazy and is is purely subjective but HDnative seems to handle VIs better.. I have not gotten any low cpu messages yet because of using VIs I will be really curious to see if I get them under the same conditions when I put tdm cards back in this computer... its no secret that tdm systems and protools in general does not play well with Vis... I use VEP to host most instruments because of that.. all of my perceived increased stability on this issue could be recoding in PT9 as opposed to it being the native aspect...I dont know

and you can chalk it up to a newer faster computer but Logic and Dp have always way out shined any protools system I owned with regard to VIs... this setup seems to be more inline with Logics Vi performance....

cheers
SP

PS I'm happy I own both but finally see the promise of a native path for those of us who have been convinced that tdm is the only viable way to do real tracking and mixing....as I first stated at least for me this thing actually works

and on another note a very special cd I recorded earlier this year comes out this week the new Roy Haynes project with Chick Corea a very fine record by a jazz giant ... if you like that sort of thing
Old 9th September 2011
  #37
WDM
Lives for gear
 
WDM's Avatar
 

Quote:
Originally Posted by scott petito View Post
No matter what computer I used on a tdm system hd3. I had 27 voices left.. I have over 90 left on the same mix native ...in addition on the tdm system I got constant dae errors not rtas errors... mix engine errors on a complex mix as I approached the limits of the cards

So thats more power huh?


Also I tracked 22 live instruments today to native with delay compensation and no low latency used full tracking... Not a peep from anyone about any latency... That s a real test for me...
It would be nice to know if you would be able to run your project on your OLD computer with HD native installed the same way as you used it with tdm cards?
Old 9th September 2011
  #38
Lives for gear
 
mykhal c's Avatar
 

Quote:
Originally Posted by scott petito View Post
yup...

and i do understand the apples to apples comparison but i don't know of any way that could truly be done... too many variables...

my point for starting the thread was to confirm that under the right conditions ie a modern computer, enough ram and a native card you can get results that rival a tdm system does it rival an hd 8 ..no... but it appears in my case to at least equal or surpass my tdm system which is an HD 3.. and at the moment that includes tracking as well as mixing..

and I may be crazy and is is purely subjective but HDnative seems to handle VIs better.. I have not gotten any low cpu messages yet because of using VIs I will be really curious to see if I get them under the same conditions when I put tdm cards back in this computer... its no secret that tdm systems and protools in general does not play well with Vis... I use VEP to host most instruments because of that.. all of my perceived increased stability on this issue could be recoding in PT9 as opposed to it being the native aspect...I dont know

and you can chalk it up to a newer faster computer but Logic and Dp have always way out shined any protools system I owned with regard to VIs... this setup seems to be more inline with Logics Vi performance....

cheers
SP

PS I'm happy I own both but finally see the promise of a native path for those of us who have been convinced that tdm is the only viable way to do real tracking and mixing....as I first stated at least for me this thing actually works

and on another note a very special cd I recorded earlier this year comes out this week the new Roy Haynes project with Chick Corea a very fine record by a jazz giant ... if you like that sort of thing
exactly Scott...too many variables. the main one bein' TDM users cannot choose what processes to offload to the cpu from the DSP cards to avoid losin' voice count. DSP cards 'were/are' the real cpu in the earlier/present days of TDM. like i said earlier, today they just cannot compete with today's rigs processing power. unfair to pit them against each other in reality...10 year old technology vs todays???!!! when the voices are gone on an i7 rig, a dual socket Xeon, whatever the rig is it is then at the mercy of the evaporated voice count. the processing doesn't just move over so the user can continue. unless one is then doin' non-related voice count processing then the show comes to a halt. and the same would be true on a Native system...it just takes a helluva lot more to get the system to that point with today's technology....a helluva lot more.

congrats on the album release too Scott. luv both of those artists!!!
Old 9th September 2011
  #39
Quote:
Originally Posted by scott petito View Post
Actually all plugins I have are both tdm and rtas except massenburg eqs a quick 135 bucks for the upgrade and the entire session came back in native with no plugin issues back to mixing 100 tracks at 88.2 with no remixing..
And my I/o setup was intact so all the outboard was as it was in tdm as well...

I'm not saying native replaces tdm I once again say that it it closer to an hd5 than an hd 3

I'm not offing my tdm setup just yet.. But the promise of a workable native system has come for me..

Cheers
SP
In that case, if you had your TDM rig in your new mac, switched all plugins on channels to RTAS and just used TDM plugins for aux channels, you would NOT be maxing out your system - and you would have more processing power spare than if you were just using HD Native.

Quote:
Originally Posted by scott petito View Post
But that's not my point my point is on a decent new mac with a native hd card I have more overall power than I do on a tdm hd3
Again only because you're comparing an older mac with a newer mac. That's not a fair comparison.

Quote:
Originally Posted by scott petito View Post
The point you are missing is on these sessions the mix engine alone is almost using two of the cards that leaves only one for processing then add the crappy voice al location that tdm forces on rtas plugs and you have fewer voices as well .. That my friend would be the same on the exact same computer and I intend to prove it as soon as my tdm cards come back from avid...
So that one card left for processing - is that not enough to run the aux channel TDM plugins? Appreciated you're running a huge session...

Quote:
Originally Posted by scott petito View Post
No matter what computer I used on a tdm system hd3. I had 27 voices left.. I have over 90 left on the same mix native ...in addition on the tdm system I got constant dae errors not rtas errors... mix engine errors on a complex mix as I approached the limits of the cards

So thats more power huh?
It would be interesting to see what would give out first though on the HD native, voices or processing power.

I do see what you're saying - for you the bottle neck is really just the voices - it's not processing power. Potentially, if you put your HD cards in a new mac, used ALL the mixer to give you as many voices as possible (or are you set like this already) and then did ALL the processing as RTAS, you may save power and get more voices. and then effectively your TDM rig would just be doing the mixing, and the rest would be native.

If this weren't a voices issue, and strictly talking about processing power, the TDM rig would always be more powerful. It's the voices restriction which is the one sticking point with TDM, and it seems that Native is just a better solution for these huge track count sessions (well - unless you went to HD4 - there will come a tipping point, be it HD4 or 5 where TDM voices aren't an issue...and then the TDM rig will be all-conquering!
Old 9th September 2011
  #40
Lives for gear
 

Quote:
Originally Posted by psycho_monkey View Post
Again only because you're comparing an older mac with a newer mac. That's not a fair comparison.
It's not a fair comparison only if you're thinking from the perspective of absolute chip count and absolute horsepower measurements.

Both you and Matt Hepworth are "technically correct" (the best kind of "correctness" right?) but it's not relevant to OPs comparison.

I think I understand what scott petito is saying and for his comparison, the correct approach is to weigh the older Mac TDM3 rig against newer mac NATIVE rig. That's actually the exact idea. He's not comparing exact host computers. He's comparing the evolution of entire rigs available to the marketplace over time. What high end complete rig was available in 2003? HD TDM & Mac G5. Ok, what high end system is available in 2011. HD NATIVE & 12-core Intel. Compare the tracking & mixing workloads on those 2 complete rigs. You couldn't buy an Intel multicore in 2003 and you can't buy a new G5 with an Apple warranty today so the host computers relevant to their respective times is part of the equation in OP's comparison.

Don't we have to go back to the philosophy behind TDM and NATIVE? Way back when (1990s), TDM guarantees a performance spec that internal Mac cpus couldn't handle. So you consider that whole "old outdated" computer augmented by TDM as "Rig-TDM". Today, you have modern computers with multicore internal cpus that can handle workloads that previously required TDM. You simply augment your Intel Mac with HD NATIVE and consider that new setup as "Rig-NATIVE". So yes, Rig-TDM and Rig-NATIVE are not identical host computers! It's not a flaw in the comparison.

To me, the OP was quite clear in post #1 that this was the observation that he was sharing. (Compare complete rigs relevant to their time periods and not raw TDM cards in identical host computers.)
Old 9th September 2011
  #41
Quote:
Originally Posted by Jason West View Post
It's not a fair comparison only if you're thinking from the perspective of absolute chip count and absolute horsepower measurements.

Both you and Matt Hepworth are "technically correct" (the best kind of "correctness" right?) but it's not relevant to OPs comparison.

I think I understand what scott petito is saying and for his comparison, the correct approach is to weigh the older Mac TDM3 rig against newer mac NATIVE rig. That's actually the exact idea. He's not comparing exact host computers. He's comparing the evolution of entire rigs available to the marketplace over time. What high end complete rig was available in 2003? HD TDM & Mac G5. Ok, what high end system is available in 2011. HD NATIVE & 12-core Intel. Compare the tracking & mixing workloads on those 2 complete rigs. You couldn't buy an Intel multicore in 2003 and you can't buy a new G5 with an Apple warranty today so the host computers relevant to their respective times is part of the equation in OP's comparison.

Don't we have to go back to the philosophy behind TDM and NATIVE? Way back when (1990s), TDM guarantees a performance spec that internal Mac cpus couldn't handle. So you consider that whole "old outdated" computer augmented by TDM as "Rig-TDM". Today, you have modern computers with multicore internal cpus that can handle workloads that previously required TDM. You simply augment your Intel Mac with HD NATIVE and consider that new setup as "Rig-NATIVE". So yes, Rig-TDM and Rig-NATIVE are not identical host computers! It's not a flaw in the comparison.

To me, the OP was quite clear in post #1 that this was the observation that he was sharing. (Compare complete rigs relevant to their time periods and not raw TDM cards in identical host computers.)
Can't say I really agree with your thinking! A tdm rig is still current technology, and you can put hd native in an older computer. The current time. All 12 of of our studio rigs run hd tdm with intel macs - are we behind the time or current?

HD didn't cone out till 2001-2 I think anyway, not 90s...pcie rigs only work in one g5 model, the current product was designed for intel macs and is still very relevant for large scale tracking...I don't disagree that tdm with g5 isn't as powerful as hd native plus intel mac, it's self-evident to anyone that's ever tried to run VIs on a g5 rig! But that doesn't alter the fact a tdm rig is more powerful than a native rig, all other things being equal.

That's not pedantry as you seem to be implying, it's true....voice count allowing as we've seen...which a bigger hd rig will deal with!
Old 9th September 2011
  #42
Lives for gear
 

Quote:
Originally Posted by psycho_monkey View Post
All 12 of of our studio rigs run hd tdm with intel macs - are we behind the time or current?
When did your facility configure and buy those 12 systems? HD Native was Announced October 2010 (don't know when it actually started shipping.) If you bought those 12 systems before October 2010 then you simply bought whatever made sense based on what Avid offered in the marketplace back then. So yes, in a certain sense, you can be behind the times only because you can't time travel to the future to buy a new product that hasn't been created yet.

Quote:
But that doesn't alter the fact a tdm rig is more powerful than a native rig, all other things being equal.
I haven't seen anyone disagree with this statement. It is correct. However, you and Matt keep ignoring that the OP was not based on the premise of "all other things being equal."

You guys want to keep the scientific rigor (pedantry?) for comparison but it's not necessary for the OP's observations.

As someone already said previously by way of analogy: people don't necessarily think of TDM +plus+ NATIVE... some think of it as TDM rig -or- NATIVE rig for a particular workload. In other words...

A horse can seat 2 people comfortably.
Maybe horses are too much of a hassle with scooping up its manure and they smell funny.
Ok, these new fangled automobiles can seat 5 people!
Therefore, a car "does more" for me than a horse.

You and Matt: "No! Your math is all a wrong. Clearly a horse and a car can be strapped together to hold a total of 7 people!!! 7 is mathematically greater than 5!"

Yes, nobody can argue that 7 is more than 5 but that's not what scott petito was saying at all. Do you disagree with this?
Old 10th September 2011
  #43
Quote:
Originally Posted by Jason West View Post
When did your facility configure and buy those 12 systems? HD Native was Announced October 2010 (don't know when it actually started shipping.) If you bought those 12 systems before October 2010 then you simply bought whatever made sense based on what Avid offered in the marketplace back then. So yes, in a certain sense, you can be behind the times only because you can't time travel to the future to buy a new product that hasn't been created yet.
No, but we've bought 2 rigs since HD Native was announced. I set the last one up on Tuesday. There's still no better system for what we do, nor for what our clients request. So...still behind the times? current product, brand new mac (well, very nearly - still came with SL thank god)...I fail to see how you think this is an out of date product, it's still the top of the Avid product line.


Quote:
Originally Posted by Jason West View Post
I haven't seen anyone disagree with this statement. It is correct. However, you and Matt keep ignoring that the OP was not based on the premise of "all other things being equal."

You guys want to keep the scientific rigor (pedantry?) for comparison but it's not necessary for the OP's observations.
I fail to see why you think it's pedantry - other than the internut arguing thing of course! OP's statement - "this rig is more powerful than my tdm rig". Response "well yeah, you've got a more powerful computer - put the cards in this one, and the TDM rig will be more powerful". Your response - "you're not allowed to do that, we have to now say that HD Native is more powerful than TDM". You've proved nothing, other than an HD rig is powerful. We knew that!

Quote:
Originally Posted by Jason West View Post
As someone already said previously by way of analogy: people don't necessarily think of TDM +plus+ NATIVE... some think of it as TDM rig -or- NATIVE rig for a particular workload.
But that's called not understanding the system properly - ignorance, if you like. Most people end up using TDM rigs in our studio with a mixture of RTAS and TDM plugins. Can't rewrite the rules to suit ignorance.

Quote:
Originally Posted by Jason West View Post
In other words...
A horse can seat 2 people comfortably.
Maybe horses are too much of a hassle with scooping up its manure and they smell funny.
Ok, these new fangled automobiles can seat 5 people!
Therefore, a car "does more" for me than a horse.

You and Matt: "No! Your math is all a wrong. Clearly a horse and a car can be strapped together to hold a total of 7 people!!! 7 is mathematically greater than 5!"

Yes, nobody can argue that 7 is more than 5 but that's not what scott petito was saying at all. Do you disagree with this?
Totally. Absolutely stupid analogy. Loads of things wrong with it - the biggest two being that a horse on it's own is in no way equal to a car, either in speed or carrying capacity (an HD rig is equal to a native rig), and that the car/horse hybrid is impractical and won't do anything of use (an HD rig in an Intel Mac is quite clearly practical AND powerful). Apart from in the extreme circumstance Scott is describing with the voice limitation (and that could be removed by adding more Process cards), adding TDM cards to an Native rig improves it in every way - strapping a horse to a car does not.

If GS has taught me anything, it's that most analogies are full of horsesh1t...if you'll pardon the pun...
Old 10th September 2011
  #44
Lives for gear
 

Quote:
Originally Posted by psycho_monkey View Post
I fail to see how you think this is an out of date product, it's still the top of the Avid product line.
Because there are 2 different (and valid) perspectives for looking at the rigs. One perspective is the absolute horsepower perspective (which you and Matt are emphasizing). The other perspective is what horsepower today at what price point today that satisfies a given session workload.


Quote:
Originally Posted by psycho_monkey View Post
Your response - "you're not allowed to do that, we have to now say that HD Native is more powerful than TDM". You've proved nothing, other than an HD rig is powerful. We knew that!
I never said that. I didn't say HD NATIVE is more powerful than TDM. Neither did the OP.

OP said he had a 1st gen Mac Pro. I assume this was a 2006 computer. He now has a new 8-core computer (2011?) with HD Native. Those are the systems he compared and they are not equal host machines. The comparison is about completing a particular session workload on 2 rigs respective of their time periods. That's the comparison and by design, the host computers can't be the same. They shouldn't be the same. It's a valid comparison.

Quote:
. Apart from in the extreme circumstance Scott is describing with the voice limitation (and that could be removed by adding more Process cards), adding TDM cards to an Native rig improves it in every way - strapping a horse to a car does not.
Well then, it's no longer HD3 ... it's now HD4 or more? Right? OP was talking about his HD3 rig. And the price would no longer be $$$, it's $$$$ more. Right? So no, adding TDM cards does not improve it in "every single way." It improves it in TDM horsepower but makes it worse in terms of cost, heat dissipation, single-points-of-failure, labor to open the machine and add the card, loss of PCIe slot that could be used for something else such as a video capture card, etc.

Neither OP nor I said all NATIVE is more powerful than any and all TDM configurations. As I said before, I violently agree with all your technical points. I'm just saying it's not relevant for this particular thread.

Please explain how all your correct counts are relevant to OP's message. Read the OP again and hopefully, you'll see what the message is.
Old 10th September 2011
  #45
Quote:
Originally Posted by Jason West View Post
Because there are 2 different (and valid) perspectives for looking at the rigs. One perspective is the absolute horsepower perspective (which you and Matt are emphasizing). The other perspective is what horsepower today at what price point today that satisfies a given session workload.




I never said that. I didn't say HD NATIVE is more powerful than TDM. Neither did the OP.
Actually, yes he did - or at least, he strongly implied it.

I quote:

"So far hd native is smoking my hd3 tdm system". No, it's your new Mac that's smoking your old Mac!


Quote:
Originally Posted by Jason West View Post
OP said he had a 1st gen Mac Pro. I assume this was a 2006 computer. He now has a new 8-core computer (2011?) with HD Native. Those are the systems he compared and they are not equal host machines. The comparison is about completing a particular session workload on 2 rigs respective of their time periods. That's the comparison and by design, the host computers can't be the same. They shouldn't be the same. It's a valid comparison.
Well, it depends if we're comparing TDM vs Native, Mac v Mac, or complete systems. What I'M saying is you can't say "HD Native smokes HD3!" because you've changed your computer.

I do see what you're saying, and why you're making a comparison, and how it relates to the OP. However it's going to be misleading for some researching this issue. And as someone who regularly gets PMs from people asking questions about PT, I'd rather have the correct information in the public forums! Not that I mind answering questions privately....it's just easier if it's all in public.

It's not about pedantry, or being right, it's about keeping information factual and not too personal. I'm personally really pleased Native is working so well for the OP, it's the only system I'd consider for my own room once the budget is in place, since I'd never have a full on tracking room.


Quote:
Originally Posted by Jason West View Post
Well then, it's no longer HD3 ... it's now HD4 or more? Right? OP was talking about his HD3 rig. And the price would no longer be $$$, it's $$$$ more. Right? So no, adding TDM cards does not improve it in "every single way." It improves it in TDM horsepower but makes it worse in terms of cost, heat dissipation, single-points-of-failure, labor to open the machine and add the card, loss of PCIe slot that could be used for something else such as a video capture card, etc.
Yes, that's all true. It's also irrelevant when it comes to how the thing actually performs when it IS working! Of course there's disadvantages to more cards - but not whilst it's working (and labour to open the machine? it took me 5 mins to rig the Mac Pro on tuesday, and that was whilst demonstrating it to a new assistant!).

Quote:
Originally Posted by Jason West View Post
Please explain how all your correct counts are relevant to OP's message. Read the OP again and hopefully, you'll see what the message is.
Hopefully I've cleared that up above. Precisely because of what you said before (people don't see Native+TDM - they just see one or the other). It's a caveat to the OP's message. The OP, whilst detailing his own valid experience, is giving a misleading view of the Native vs TDM argument - which leads to people like me having to explain to people how things actually are. I'm basically selfishly trying to keep the facts in the public domain to save myself work later on
Old 10th September 2011
  #46
Lives for gear
 
mykhal c's Avatar
 

i'll say it again in a different way...as long as a TDM (HD3 user for example) can live harmoniously within his voice count, regardless (mostly) of his cpu's power, he/she will probably be/stay/remain happy. as long as HD Native users usin' today's Intel (mostly) are not concerned with a session exceedin' 64 simultaneous I/O they'll remain happy. and Scott is happy!!!

EDIT: and to keep it real...for example...192 track session, loads of plugs, delay comp, etc etc etc runnin' on an i7 or dual Xeon (take your pick) is NOW ported to an i7 or dual Xeon (take your pick) usin' HD3 WILL FLAT OUT BRING IT TO ITS' KNEES!!! END OF STORY!!! so Scott is right...it will smoke his HD3. he's proved it..regardless of the fact that some think it's because of his old Mac...not so. that's just the truth...for real.
Old 10th September 2011
  #47
Lives for gear
Quote:
Originally Posted by tobymusic View Post
Great to hear it's working out good for you. I've been not so happy, on my 2008 8-core I'm getting a much inferior performance compared to running the TDM mixer (and a few TDM plugs) on my HD3. The mixer and automation really slows everything down - not convincing. It might actually be an incompatability between HD Native and the 2008 Mac Pro machines.

T.
Sorry to back track the thread a bit but I recently switched over to an HD native / 192io / 2008 8 core Mac Pro system as well and the performance has not been great.

My system usage runs at 6% with no tracks and no plugins. With aprox. 30 tracks and maybe 25 plugs going I'll start to get errors with a buffer of 128 and sometimes even at 256.

I know it's not the fastest Mac out there but every story I hear is about people running huge mixes at a 64 buffer. The recent tape op reviewer said he had an 87 track "dense" mix and his CPU usage was only at 5%! Something has to be off here...

I keep a very clean system, optimized, with legit and fully up to date plugins. I've owned, and used professionally every Pro Tool system available since my nubus Pro Tools 3 system in 1998 so please no "are you recording to your system drive" questions.

Any help is greatly appreciated.

P.S. I'm on 10.6.7 and PT hd 9.0.3
Old 10th September 2011
  #48
Lives for gear
 
mykhal c's Avatar
 

Quote:
Originally Posted by SaschaP View Post
Sorry to back track the thread a bit but I recently switched over to an HD native / 192io / 2008 8 core Mac Pro system as well and the performance has not been great.

My system usage runs at 6% with no tracks and no plugins. With aprox. 30 tracks and maybe 25 plugs going I'll start to get errors with a buffer of 128 and sometimes even at 256.

I know it's not the fastest Mac out there but every story I hear is about people running huge mixes at a 64 buffer. The recent tape op reviewer said he had an 87 track "dense" mix and his CPU usage was only at 5%! Something has to be off here...

I keep a very clean system, optimized, with legit and fully up to date plugins. I've owned, and used professionally every Pro Tool system available since my nubus Pro Tools 3 system in 1998 so please no "are you recording to your system drive" questions.

Any help is greatly appreciated.

P.S. I'm on 10.6.7 and PT hd 9.0.3
don't wanna thread steal but look at your ram. i'm no Mac guru but have built and supported a gazillion PC/Windows builds and i always look at ram. try pullin' out all sticks except one and test your session. if you see improvement re-insert 1 dimm at a time until everythin' is back. hopefully it'll help you locate the problem dimm...if it points to ram. hope that helps. back to regularly scheduled programming
Old 10th September 2011
  #49
Lives for gear
Thanks mykhal but I've already done that and my ram seems fine. I should also point out that I had been running a FW based 003 system prior to the upgrade without issue and with better performance.
Old 10th September 2011
  #50
Lives for gear
 
scott petito's Avatar
 

[QUOTE=psycho_monkey;7016404]Actually, yes he did - or at least, he strongly implied it.

I quote:

"So far hd native is smoking my hd3

Ok this is the part of my post that is the problem I should have been much more clear... I absolutely think about tdm and rtas or native if you like working in tandem and in fact that is the only way I could have finished some of my past projects....

I am suggesting or implying that tdm may be creating some of the rtas logjams by increasing voice count use and therefore also robbing the hd cards for other duties like running more plugins....
Also check out the delay compensation on a track in tdm mixing rtas and tdm the comp hit is higher! Than the same plugging report in native system this should have nothing to do with which mac we talking about... And is a mystery to me unless there is something in the tdm coding that is causing this

Delay compensation on PT hd has also had a long standing well documented bug that turns it off randomly. This has so far not been the case on my native system....
I guess if I really want to make you guys mad what I'm coming to think is that tdm is actually crippling the use of native ! It may be better to live in one world or the other and in fact I know many guys who refuse to mix rtas and tdm plugs for this reason....

Ah what do I know ! It did pay for it self in about 4 days !

Cheers
Scott
Old 10th September 2011
  #51
Quote:
Originally Posted by scott petito View Post
No matter what computer I used on a tdm system hd3. I had 27 voices left.. I have over 90 left on the same mix native ...in addition on the tdm system I got constant dae errors not rtas errors... mix engine errors on a complex mix as I approached the limits of the cards

So thats more power huh?


Also I tracked 22 live instruments today to native with delay compensation and no low latency used full tracking... Not a peep from anyone about any latency... That s a real test for me...
I assume you played with the voice allocation settings on the TDM hardware?
Assuming that you've chosen to use RTAS plugins (same as your Native system), you should be able to
configure for more voices on the TDM hardware ( and less DSP for plug-ins).

In that regard, an HD-3 system (dialed in for more voices and less DSP) should be able to do what a Native system can do.

jeff
Old 10th September 2011
  #52
Lives for gear
 
KBOY's Avatar
 

But you've never ran your TDM system on the new mac. Which is totally confusing me. I can't figure out how you came to the conclusion of more cpu usage in regards to latency compensation.

I am thinking like others in this thread that your new mac smokes your old mac.
Old 10th September 2011
  #53
Doesn't make a whole lot of sense to me either .. the MIX engine is running on an
FPGA on the Native card. The rest of the processing is done in RTAS ... which will
undoubtedly have latency for the plugins that are used in-track.

Given a very small buffer size (64 bytes for example), the latency will
be short .. ( on the order of 1.33 ms at 48kHz). That's assuming there's no
ADC required for plug ins that don't process in realtime.

If there are no plug ins on any of the tracks, busses or master faders, then
the Native system will perform as a TDM system does.

If there are plug-ins .. it will most certainly not be free of latency. It
may be quite short (given the buffer size).

One benefit of the Native Mix engine is usage of double precision
floating point ...

jeff
Old 10th September 2011
  #54
Quote:
Originally Posted by mykhal c View Post
EDIT: and to keep it real...for example...192 track session, loads of plugs, delay comp, etc etc etc runnin' on an i7 or dual Xeon (take your pick) is NOW ported to an i7 or dual Xeon (take your pick) usin' HD3 WILL FLAT OUT BRING IT TO ITS' KNEES!!! END OF STORY!!! so Scott is right...it will smoke his HD3. he's proved it..regardless of the fact that some think it's because of his old Mac...not so. that's just the truth...for real.
I think that sort of session would bring most rigs to their knees - multiple HDs would be needed for a start. The difference is that an HD rig would either work or it wouldn't (I've never seen this whole instability thing when approaching limits - obviously I've seen the re-allocation thing which is annoying, but I've always had rigs work or not allow any more plugins). A native rig would start to stutter. So no - unless you can offer examples, I don't think you can say this - just shouting doesn't make it "end of story"!

Quote:
Originally Posted by scott petito View Post
Ok this is the part of my post that is the problem I should have been much more clear... I absolutely think about tdm and rtas or native if you like working in tandem and in fact that is the only way I could have finished some of my past projects....

I am suggesting or implying that tdm may be creating some of the rtas logjams by increasing voice count use and therefore also robbing the hd cards for other duties like running more plugins....
Also check out the delay compensation on a track in tdm mixing rtas and tdm the comp hit is higher! Than the same plugging report in native system this should have nothing to do with which mac we talking about... And is a mystery to me unless there is something in the tdm coding that is causing this
I've never seen this, and I think jmarkham has put my viewpoint already.

Quote:
Originally Posted by scott petito View Post
Delay compensation on PT hd has also had a long standing well documented bug that turns it off randomly. This has so far not been the case on my native system....
I guess if I really want to make you guys mad what I'm coming to think is that tdm is actually crippling the use of native ! It may be better to live in one world or the other and in fact I know many guys who refuse to mix rtas and tdm plugs for this reason....

Ah what do I know ! It did pay for it self in about 4 days !

Cheers
Scott
Great news on the paying for itself - at the end of the day, it's clearly the right product for you.

Guys who refuse to mix TDM and RTAS....well, I hope they're good engineers because they're clearly not programmers! Is this like the poster who insisted leaving a gap between plugins made things sound better?

I've also seen a delay compensation bug in regular PT9, but I put it down to the lynx FW driver - always seemed to occur after needing to restart PT because of a FW drive being connected and the lynx freaking out.
Old 10th September 2011
  #55
Lives for gear
 
zak7's Avatar
 

Quote:
Originally Posted by psycho_monkey View Post
Well, in that case, assuming your TDM plugins have RTAS equivalents, your native rig runs plugins AND the mixer, and your TDM rig just runs the (RTAS) plugins, leaving the mixer to the DSP cards. Consequently, your computer still has power free.

Matt IS correct - not saying all will need it, but assuming the same computer power (which is NOT the case with the OP) a TDM rig always has TDM power IN ADDITION to what the computer can provide, so with ALWAYS be more powerful.

A couple of caveats - if you're running RTAS plugins with no TDM equivalent, particularly on aux channels, you may find yourself running out of voices, to the point where you've used up so much TDM power for the mixer the improvement is small (but there will still be an improvement). Likewise with virtual instruments, the routing may well use up extra voices.

The main difference is that Native is an option for those who don't NEED the extra power or who don't do large scale tracking. It's not going to replace an HD3 rig for those who really need one - but it'd be better than an HD1 in just about every scenario.
get your point, so basically what you are saying is that if you have a HD3 TDM using only RTAS and then a HD Native ...where in both escenarios u have same Mac Pro 8 Core....you you are saying that the TDM system is more powerfull because take care of the Tracks?? well, is not also that the TDM system will limit your voices ?

How many tracks is the limit with and HD Native??
Old 10th September 2011
  #56
Lives for gear
 
scott petito's Avatar
 

Found this
Place a TDM plugin in the first insert, and the audio must use a Voice to get into the TDM Environment, and then once there, it will flow to the proper DSP chip(s) for the TDM plug-in processing. If you place an RTAS plugin in any insert slot AFTER that TDM plug-in, what will happen? Well, the audio must enter first, do the TDM process, and then exit, receive RTAS processing, and then enter once again. That means you need three of those "one way, single lane portals" for each track width of audio. For ten mono audio tracks that have a TDM plug-in in the first insert, followed by 3 RTAS plugins in insert slots 2,3,and 4...well, you have ten lanes. And you have to enter for the TDM process for each. That is Ten Voices so far. But, you now need to exit for the RTAS processes. Once out, all 3 of those RTAS processes can be done in the host BEFORE returning to the TDM Enironment. Getting back in will require a Voice for each track width. If only mono plug-ins were used, then it is: ten voices to get in, do TDM processes, and then ten voices to get back out to do al the RTAS stuff, and then ten more to get back in once again. 30 Voices total.

Say you used a Mono-Stereo TDM plugin on each of those ten mono tracks. To get in, it is still only a mono signal entering. So, ten voices. But getting back out? Well, the signal width has been ncreased from 1 to 2. So, you now need 20 voices for them to exit for the RTAS processing. And another 20 to get back in. Total for those ten tracks of mono audio, that receive mono->stereo TDM processing, and then stereo RTAS processing? 50 voices.

My new native system is still using only ten voices.. Regardless of computer
That's how I see tdm in a sense crippling the native implementation

Do it at 96k then look at your tdm voice count

Cheers
Scott
Old 10th September 2011
  #57
Lives for gear
 
mykhal c's Avatar
 

actually wasn't thinkin' of it as shoutin'...sorry...my bad. my point in my example really was you can't even begin a session like this on HD3. the mere fact that there are 192 tracks involved stops it right there on an HD3 system...you can go no further regardless of havin' a super i7/dual Xeon or not. this is the flaw in the old Mac argument people aim at Scott. on HD Native i could 'continue' to build this 'template' if you will. i could throw plugs on tracks and hit play at a 1024 buffer (probably 512 or even 256 dependin' on rig) and it would play. and yes i know there's no audio streamin' at this point. but the template example also makes my point. try this on your HD3 systems and then go to your local Avid dealer with HD Native set up and ask them to draw up this exact template. you'll see proof and will 'have' to acknowledge things then. until then, Scott 'has' posted his 'real' world findings. lotta truth that brought to the discussion, huh?

EDIT: i've stated on other forums/discussions, as my best guess, that in the future Avid will unveil new super I/O cards for Native...128, 192, 256. this addition along with the new HD I/O interface places Avid right where they wanna be...lettin' this and future generations of super computers do the work that the DSP cards used to (and still provide) have to do when they were unveiled. just 'my' guess. but times 'have' changed and Avid has shown you their direction...believe it or not.
Old 11th September 2011
  #58
Lives for gear
 
scott petito's Avatar
 

It would be nice to know if you would be able to run your project on your OLD computer with HD native installed the same way as you used it with tdm cards?



Im am quite certain I could not get anywhere near the performance on my now 4 plus year old mac using Native HD...

but that is also why Avid waited until a fast computer was available that could be stable and dependable( like an HD system) to release a professional native solution....

and I will now in front of everyone correct my "smokes" statement and say:

My new native system running on my new mac (total cost about 6.5 grand )

smokes my TDM HD3 system running on my old mac (total cost new was about 17 grand) not to mention the then double or triple cost for TDM plugins over RTAS...

and more proof to me was today I opened a file on my native system recorded at real world at 88.2K on a HD7 with only TDM plugins

I could not open this file on my HD3 last month I had to create a new session and import the audio at 44.1K

any questions?


cheers
Scott

by the way I hate the fact that I spending so much time shilling for Avid but they really got this right in my opinion the caveat being you need a new fast mac!
Old 11th September 2011
  #59
When you say you couldn't open it, what exactly happened? PT
should have disabled the plug-ins once the DSP had been exceeded
and disabled tracks as well.

I've not had a session that point-blank wouldn't open in some time.

jeff
Old 11th September 2011
  #60
Here for the gear
 

“ and more proof to me was today I opened a file on my native system recorded at real world at 88.2K on a HD7 with only TDM plugins

I could not open this file on my HD3 last month I had to create a new session and import the audio at 44.1K

any questions?”
Yes-
What would happen if before you opened the session on an HD3 on a new mac pro you changed enough of the TDM plug ins to RTAS format that only 3 cards worth of TDM was being used? - Other than flexibility this most recent example doesn’t illustrate the advantages of one system over another.
If a session is in danger of maxing out due to voices native could have advantages. Many sessions never are close to maxxing out voices but are close to maxing out dsp power, in which case a TDM system with an equally powered computer could be more powerful.
Regarding voices, if having a TDM plugin after an RTAS eats up voices, if all your plugins are both TDM and RTAS simply convert the TDM plugins which follow to RTAS and on audio tracks your Native vs TDM voice count is the same.
Market price medium level new mac pro+ used native card <$6k>
Market price HD3 with the same new mac pro plus new Magma Chassis ($2.2K) + used PCI-x core+2accel ($2k) <$8k>, HD6 then about $9.5k). TDM versions of plugins can drive up the cost.

If you receive a session with any of the following plugins in TDM they will not open up with their settings automatically in a Native session.

Acess Virus Indigo
Amp Farm 3.1
Antares/Microphone Modeler
Aphex Aural Exciter/Big Bottom
Avid Heat
Avid/Reverb One
Bruno
Cranesong/Phoenix
Drawmer/Drawmer Dynamics
Eventide/Anthology II Bundle
Focusrite/Liquid Mix HD
Impact
Line 6/Ampfarm
No-noise
Reso
Reverb One
Revibe
Soundtoys/Soundblender
Soundtoys Timeblender
TC/DVR 2
TC/Harmony 4
TC/Master S
TC/MD 3
TC/Non Lin 2
TC/Tube Tech CL 1b
URS Double Precision EQ's
Indigo
Serato Rane series Dynamics, Compressor and gate
Waves PS22
eventide factory
eventide octravox
eventide 949
TL space TDM (long/medium/short)
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump