The No.1 Website for Pro Audio
Apogee/Symphony Performance Compared to Protools HD
Old 3rd February 2007
  #91
Gear Maniac
 

Quote:
Originally Posted by Threshold View Post
nikki-k, & everybody else,

It's like cancer. Nobody wants it once diagnosed.

Threshold. stike
DO people want it before diagnosed?

Seriously man, I understand your desire for accurate information...I think that's great. I just don't get why you would ask a bunch of guys on an internet forum for information, when getting information from a bunch of guys on an internet forum isn't really what you want, ya know what I mean?

"Hey,

I thought delay compensation in Logic Pro only worked with Logic Pro plugins, & not audio units, or outside vendor plugs. Correct me if I'm wrong."

I know this to be an incorrect statement. A few guys chimed in ...with a spirit of helping you out and giving you info. Now, I don't think you intended your response to ilicit this, but when someone tries to answer your question, your responses come off reading "Your answer isn't good enough for me...you don't know what you're talking about. Go do the legwork that I should have done myself and find accurate rescources from people who are smarter than you!"

Like I said, I don't think that was your intention, but it's hard not to take it that way when trying to offer help to someone that has asked for it, and then refuses it because of who's offering the help.

If reports from manufacturers are what you are seeking, you may want to ask for that specifically ahead of time, or go do the legwork yourself.
Old 3rd February 2007
  #92
Gear Nut
 

synthdogg, & all,

No offense intended to anyone. And, those who might have taken offense should'nt take such. This reply would'nt be coming from me if the questions replied to in this forum, & others were consistently accurate. Nikki-k's previous post backs up my position. These forums 'can' get out of hand.

Forums like this are good for information if the data is solid. Many times I've seen people on this site give reference information, or links therefore the request them from me. And, I do do the research myself. If you check my post history I've provided info for people with data I can 'verify' most times. Give good advice/suggestions, & you should expect to receive the same back. That's part of my motto for maintaining around here. Also I don't own every singel piece of equipment mentioned here. If I did I'd call the respective manufacturer for technical assistance. After all what are we here for?

Threshold.stike
Old 3rd February 2007
  #93
Lives for gear
 
nikki-k's Avatar
 

Quote:
Originally Posted by Threshold View Post
nikki-k, & everybody else,

.......Sales are up, and there are still a large number of people who realize that PT HD offers more than simple "low latency while tracking" benefits. It is a different approach than native...but, I outlined this in my previous post (which was probably ignored). Funny thing is, as a PT HD user, and one that loves PT software, I still very proudly claim it is not for everyone, and that native solutions are excellent, and that it is so very cool we have so much to choose from. The only real sad part I see is the constant misinformation that gets circulated, read, regurgitated and then spewed forth amid a flow of extra %&@+^ that would boggle the mind if it were not so apparent how flawed humans were at the core to begin with.

It's funny- all it takes is one person with an audience to proclaim it is ridiculous to even think of spending money on sneakers to climb Mt Everest, and you have a huge following proclaiming the same, with no proof or reason to do so other than the simple acceptance of what the first moron claimed. ........


-nikki-k



This is exactly why I have been asking people earlier in this thread to provide reference specifications/links to confirm data in a 'responsible' manner, & not communicate information just based upon personal experience which can be biased, skewed, or flat out inaccurate because the person conveying the opinion just flat out does'nt know what in the audio universe they are talking about(my reason for being so adamant about facts relating to Logic Pro's delay compensation for non Logic Pro Audio Units plugins, & etc.). I want the 'truth'. Also I talk to several other audio pro's who respect my opinion, & I theirs. I don't like, or won't repeat things that are questionable to others. It's like cancer. Nobody wants it once diagnosed.

Threshold. stike
I am hoping you were not using that post of mine as an example of what you DONT want in this thread. If so, you took it so severely out of context that it actually provides yet another shining example of what I was saying when I added that to a set of FACTS.

Since the responses to the original post drifted and varied, and although my first response in this thread was one of providing facts that were not being properly conveyed, I saw that despite what I thought was clear explanation, people STILL persisted in presenting untrue information, accompanied by opinions of preference (which is what I thought your reply was referring to Threshold! Hope I am right there...lol...). Any non-factual comments were made simply to further try to push the thread back to it's intent: discovery of facts, with specific reference to two choices. In that this did not specifically ask for the broadbased facts I attempted to provide, the very nature of the question and subsequent responses seemed to show a lack of understanding those basics, which then would almost demand that they be presented, or at least provided for any who lacked them.

The facts are simple:

native solution, live tracking, can be done with only latency of converters. Add plug-ins, and you HAVE TO ADD the buffer being used. This is LAW. This is FACT. The Apogee solution being presented provides a strong(er) set of drivers that people are claiming will allow the user to utilize lower buffer settings, thus acheiving a lower total latency, or "trip time" for a signal, when placing plug-ins inline, and when confronted with an existing session/project that requires live overdubbing to be done.

PT HD, when used with TDM plug-ins, provides the same thing, but there is no buffer to contend with, and plug-in process times will ALWAYS be added to the signals "trip time." In many cases the process time for a TDM plug-in or plug-ins will be as low as one sample or maybe several samples. This would mean the native solution could have about 30 samples or so more latency than PT HD, or much more. I think most would agree that 30 samples of difference is quite low, and almost not worth worrying about. Work with heavier session/project loads in a native enviornment, and that "~30 samples" number increases, typically doubling or more as load increases and a higher buffer is required. As to how this difference affects one's perception of this delay...well, that is completely and fully an individual thing. And once again, only has direct pertinence when involving the monitoring of a live signal while tracking/overdubbing.

Inasmuch as the original post also inquired of the reality of Apogee's cliams for their native solution...well, I never provided any info on that, as I have not used the system. It looks good on paper, and I have heard people report good results. The PT HD numbers and facts I presented can be verified with several minutes spent on a properly configured system. I would gather that due to the nature of how session variables would affect a native based session/project, verifying the Apogee numbers might take just a little more time on a system porperly configured. Regardless, once the base facts are outlined and understood, I would trust only my own fact collecting experiences before a final decision and purchase.

(maybe it is time to reconsider forum browsing on the internet...it almost isnt worth the frustration to try and help people )
(oh- an dthis is what I really dont like about writing responses vs talking to someone face-to-face; the true intent can so easily be lost...)
Old 4th February 2007
  #94
Lives for gear
 
nativeaudio's Avatar
 

Quote:
Originally Posted by nikki-k View Post
Just as the latency quoted for the Apogee stuff running natively depends upon the amount of DSP power you purchased (ie, top end Mac Pro vs older G4).
I'm sure you'll need a MacPro to get the optimum/reference performance. Saying that you'd get the same performance from a G4 or a G5 would be misleading.

Quote:
PT HD can be expensive, but it is all relative. If you have a business that can benefit from using an enclosed enviornment (PT HD, Pyramix, PARIS, etc) over a native one, and business is good, then I would assert that cost differences here are not nearly as important as some will assume.
Sure, but PT HD vs. Symphony is more about choosing a system you want to invest in, (AD16 and DA16 aren't the cheapest I/Os out there, as we know), more than finding a way to save $5000 only.

Quote:
Hmmm...interesting thing...What is the buffer setting they have set for getting those 200 instances?
They mention 32 samples on that page, I don't know if this is valid for all their tests.


Quote:
And how about simple math? If one had PT HD with that situation, those 200 plug-ins would still be available...along with 35 more instances running on the HD Cards.
Sure.
Quote:
Oh...and the user would get the first 35 of those "free."
Depending on the definition of free, those 35 instances would be the least free plug-ins in his setup!
Old 4th February 2007
  #95
Lives for gear
 
nativeaudio's Avatar
 

Quote:
Originally Posted by nikki-k View Post
In many cases the process time for a TDM plug-in or plug-ins will be as low as one sample or maybe several samples. This would mean the native solution could have about 30 samples or so more latency than PT HD, or much more.
Hi, what do you mean when you say that a native solution could have about 30 samples or so more latency than PT HD, or much more? I know someone who added an EQ and a comp to the recording chain on a Mac Pro and a Symphony setup, measured the result, and the plugins didn't add any latency. Zero samples, which is not possible in a TDM rig. Not that it matters, because 2 samples @ 44.1 = only 0,05 milliseconds, so a normal TDM EQ and TDM comp won't add any real latency anyway, but I just didn't understand what you were trying to say. There's no reason I know of to assume that there would be more latency running the plug-in itself. (And by the way, a native plugin running on a MacPro runs on chips 10 times as fast as the TDM chips). A more significant risk, however, on a native system, is that you at some point would have to go from the 32 buffer the 64 buffer - is that what you're referring to? This would probably add between 1 and 2 ms of latency compared with the 32 buffer setting. Then you'd end up around 5 ms @ 44.1, circa twice as much as you'd have on a PT HD rig. You could of course freeze tracks (easy, and probably more flexible in future versions of Logic), or use another Mac as a node (like a cheap Mac Mini), but then the whole thing is more about is where workflow/price etc.

Some problems with TDM rigs is that you can't bounce offline, use Digidesign's interleaved file format, you still need to make two worlds 'talk to each other' (TDM-chips and the Mac CPU), they are less portable, you are dependent on Digidesign's price and upgrade policy/prices and more, but they also have their benefits - but these things are covered in other threads.
Old 4th February 2007
  #96
Lives for gear
 
nativeaudio's Avatar
 

Quote:
Originally Posted by nikki-k View Post
Not entirely true...

In any world, certain plug-ins require "X" more time than ANY buffer size due to the type of process it is performing. As an example, a compressor might use a method to look-ahead slightly
Sure, and that's what I meant by mentioning especially demanding plug-ins (spelled algo in that other post)... :D
Old 4th February 2007
  #97
Lives for gear
 
nativeaudio's Avatar
 

Quote:
Originally Posted by synthdogg View Post
Testing at 44.1 yielded 3.3msec of latency, analog to analog.
That's way better than the *real life* roundtrip latency results from other native rigs posted over the last few years (128 samples X 3 + two 'normal' converters = easily 10 milliseconds), or 256 samples X 3 plus converters = circa 20 milliseconds.

Maybe the new 'high performance' drivers are even faster, but even if they're not, we're still in the less than one milliseconds away from the PT HD range. According to Digidesign, the PT HD latency is 105 samples. That's 2.38 milliseconds @ 44.1kHz, or 1.09 ms @ 96.

At 44.1, the difference is 0,92 ms between the two systems, or 0,54 @ 96 kHz. Or, as John Lennon would have said it, The War Is Over.
Old 4th February 2007
  #98
Lives for gear
 
midnightsun's Avatar
 

Quote:
Originally Posted by nativeaudio View Post
Hi midnightsun, can you add a normal EQ and a normal compressor to a recording track, measure the difference with/without these plug-ins, and post the results? It would be very interesting to hear what you find.
But, first tell me exactly how you define latency and exactly what values I should measure. (A)-- Do you want me to measure the time it takes for an audio signal to hit the AD converter, go through Logic Pro (with and without plugins), DA, and out? Or (B)-- are you interested in the determining a difference in monitoring delay between a recorded track without plugins and with plugins? I have already done this test and can tell you than with (A) adding plugins will increase the time delay. With (B) there would be no differernce between the measurements with and without plugins. I have no experience knowing if non Logic plugins hold consistent. If time permits and you wish I can repeat the study and carefully log the exact numbers for reporting.
Old 4th February 2007
  #99
Lives for gear
 
nikki-k's Avatar
 

Quote:
Originally Posted by nativeaudio View Post
Hi, what do you mean when you say that a native solution could have about 30 samples or so more latency than PT HD, or much more? I know someone who added an EQ and a comp to the recording chain on a Mac Pro and a Symphony setup, measured the result, and the plugins didn't add any latency. Zero samples, which is not possible in a TDM rig.
OK...use plug-ins on a native rig, and you have to incur the penalty of the buffer. No way to work around this, no way to bypass it- it is LAW. If I add two plug-ins that are TDM in PT HD that each take one sample of process time, then it is converters + TDM 1 + TDM 2. In native, the equation is converters + buffer. Since converters can be the same in both equations, they cancel. Since I established the plugins as each being 1 sample apiece for TDM that I chose, and since th elowest you get with SYmphony is 32 samples, we now know the difference is 30 samples, against the native solution. If the session/project was large, and had a bunch of plug-ins already, pushing the limits of the system @ 32 sample buffer, then they might have to jump to a 64 sample buffer. Thus, could be double, or more.

With no plug-ins being used with the Symphony solution, or many other native solutions, a direct patching scheme can be used, such as ASIO2 provides for, and such as RME has wth their cards/drivers. Thus, only the converters' process time without plug-ins...and PT HD would be the same. Add these two, and you derive what I posted.

Quote:
Originally Posted by nativeaudio View Post
Not that it matters, because 2 samples @ 44.1 = only 0,05 milliseconds, so a normal TDM EQ and TDM comp won't add any real latency anyway, but I just didn't understand what you were trying to say. , There's no real reason I know of to assume that there would be more latency running the plug-in itself. By the way, a native plugin running on a MacPro runs on chips 10 times as fast as the TDM chips). A more significant risk, however, on a native system, is that you at some point would have to go from the 32 buffer the 64 buffer - is that what you're referring to? This would probably add between 1 and 2 ms of latency compared with the 32 buffer setting. Then you'd end up around 5 ms @ 44.1, circa twice as much as you'd have on a PT HD rig. You could of course freeze tracks (easy, and probably more flexible in future versions of Logic), or use another Mac as a node (like a cheap Mac Mini), but then the whole thing is more about is where workflow/price etc.
In pure factual terms, yes the 30 sample difference I mentioned is valid...but, once again, it comes down to what YOU are happy with. If that 30 sample difference is too much, then PT HD might be a better choice. If you forsee the possiblility of greater buffer sizes being needed, despite the amount of native power we have, then once again- PT HD. But, for many people, that 30 sample (potential) difference is nothing. And the difference in price, and requirement to use a DAW they might not care for makes the native solution VERY attractive, and very cool to have.

See, this is what it comes down to. Both solutions are powerful, and both can do the jobs they claim. It is OUR job to simply become as educated as possible, and then select the one (or more!) that we feel most comfortable on. I have a feeling the original poster was simply looking for a yes or no as to the claims made by Apogee in reference to Symphony solution.

Quote:
Originally Posted by nativeaudio View Post
Some problems with TDM rigs is that you can't bounce offline, user interleaved files, you still need to make two worlds 'talk to each other' (TDM-chips and the Mac CPU), they are less portable, you are dependent on Digidesign's price and upgrade policy/prices and more, but they also have their benefits - but these things are covered in other threads.
Exactly. Strictly speaking in factual terms, relating to Symphony and PT HD, it is laid out, and strictly a matter of discovering which systems are capable of what: does one of the available Mac Pro's allow for the performance claimed by Apogee? If yes, then there ya go! From there, the potential buyer can weigh thos eother factors, and then decide.

And yeah- Digi is a love/hate relationship, and it is the same with Microsoft, Apple, Gibson, Sony, etc, etc ,etc...lol!!!
Old 4th February 2007
  #100
Lives for gear
 
nativeaudio's Avatar
 

Quote:
Originally Posted by midnightsun
-- Do you want me to measure the time it takes for an audio signal to hit the AD converter, go through Logic Pro (with and without plugins), DA, and out?
Yes, that's the interesting part, with Software Monitoring ON in both cases...




Quote:
Originally Posted by nikki-k View Post
Since converters can be the same in both equations, they cancel.
That's true WHEN they are the same. I suspect that the Apogee converters are faster than the 192 converters (more maths below).

A PT HD rig with a 192I/O has a 105 samples roundtrip @ 44.1. (A Mix rig uses less, but nobody noticed that HD systems were 'slower' - even after they had been told that they were, nobody cared, because 0.4-0.5 milliseconds isn't very much.)

A Symphony rig with AD16 and DA16 @ 44.1 seem to use 145 samples @ 32 buffer. 145 minus two buffers (145-64) = 81 samples. But there's more: the system. OSX is a multiclient OS, and this takes some overhead. NB: this overhead is included in these 81 samples.

Quote:
Since I established the plugins as each being 1 sample apiece for TDM that I chose, and since th elowest you get with SYmphony is 32 samples, we now know the difference is 30 samples, against the native solution.
It seems that you forget the system overhead: in a native rig, the audio signal travels into the OS and out again. From what I heard a while ago, this does not happen within the buffer size latency. Do we know for sure that there's no overhead baked into those 105 samples PT HD uses either, or if these 105 samples are only AD+DA? Not that long ago, Digidesign sold mBox I/Os meant for the native market, locked to go no lower than 512 samples buffers (times two): that's at least 23 (!) ms latency, not counting the converters. Some TDM users bought that as an addition to their PT rig and concluded that 'native stinks'.

Quote:
If the session/project was large, and had a bunch of plug-ins already, pushing the limits of the system @ 32 sample buffer, then they might have to jump to a 64 sample buffer. Thus, could be double, or more.
True. ANd if a faster Mac comes out tomorrow, allowing for use of a 16 sample buffer setting, the total roundtrip would be lower than today. Will faster Macs be released later this year? Yes. Will Apple release Macs next year that are faster than the 2007 models? Yes.

Quote:
With no plug-ins being used with the Symphony solution, or many other native solutions, a direct patching scheme can be used, such as ASIO2 provides for, and such as RME has wth their cards/drivers. Thus, only the converters' process time without plug-ins...and PT HD would be the same.
This is correct, if they able to use - and are using - equally fast converters with no added in-between-stuff that takes time, and there's no overhead baked into the mentioned 105 samples from Digidesign. AD+DA on a 192I/0 takes 2,38 ms (if there is no overhead or other time consuming stuff), meaning circa 1.2 ms pr. converter. The AD16/DA16 must use less, because they use eg. 1.6 ms - 154 samples-64 samples (buffers) = 90 samples = 0.93 ms including system overhead. That's 90 samples for DA + AD + system overhead. It seems that the best way to get low latency on a PT HD rig is to use Apogee converters.

Right now, what we know this winter, is that the difference for the first time is very small, and slightly lower on a PT HD rig.

Quote:
In pure factual terms, yes the 30 sample difference I mentioned is valid...but, once again, it comes down to what YOU are happy with.
Sure, and to clarify: The difference isn't that using one plug-in on a native system adds 30 samples, but that not using no-latency-mode on a native rig (Software Monitoring OFF in Logic) adds some extra latency. Unlike on a PT HD rig, a native rig allows users to bypass the 'audio brain', kind of, meaning even lower latency than we talk about here (but no plug-ins on the recording track).

So, only to avoid confusion, adding two plug-ins would NOT mean adding 60 samples. Adding one 'normal' plug-in in a native rig adds zero samples actually, as explained above, and as confirmed by tests (not by midnightsun's test). But as you say, Nikki, the difference between a Symphony rig and a PT HD rig right now is circa 30 samples, or to be more correct: circa 40. (105 samples vs. circa 145 samples). Actually, maybe even 49 samples (1.6ms @ 96k = 154 samples = 49 more than Digidesigns 105 samples, if these numbers are still correct, and also correct @ 96k). I'd like to see more detailed info from both Apogee and Digidesign, not that plus or minus eg. 10 samples by itself matters much in most situations - we're talking about 0,2 milliseconds or less here - it's just my inner nerd that wants some nourishment.

It's all about which system(s) you want to invest in for the future. I've decided to go the native way, and have friends who have done the same, while others are still on TDM. Good for me, because then I can visit their studios if I need to open one of my hundreds of old TDM sessions to check which plug-in settings I used in my Logic+TDM songs.


[Edit: I just came across a post from Apogee Max I had seen earlier but forgotten about, saying that the latency roundtrip @ 44.1 = 3 ms. If this is the case, and not only a roundoff, the difference is 27 samples; even less than I described above]
Old 4th February 2007
  #101
Lives for gear
 
barryjohns's Avatar
 

nativeaudio, I am so glad you joined this topic. You have clarified so much. I appreciate your taking the time to share your knowledge.
Old 4th February 2007
  #102
Gear Maniac
 

I'm a hardened logic user currently using ptmix3 - I like the protools plugs and it's low latency. I'm upgrading to a Mac Pro dual 3 and like most will have to wait until Apple get their arse in gear and allow TDM to run on Logic via the ESB (version 8 fingers crossed). I'm thinking a good interim might be to get the apogee DA16x (maybe with the symphony) and a firewire TC and PCI-E UAD. Or get the HD-X card and just wait. Any opinions on the best solution?
Old 4th February 2007
  #103
Lives for gear
 
nikki-k's Avatar
 

Quote:
Originally Posted by nativeaudio View Post
Quote:
Originally Posted by nikki-k
Since converters can be the same in both equations, they cancel.
That's true WHEN they are the same. I suspect that the Apogee converters are faster than the 192 converters (more maths below).
A PT HD rig with a 192I/O has a 105 samples roundtrip @ 44.1. (A Mix rig uses less, but nobody noticed that HD systems were 'slower' - even after they had been told that they were, nobody cared, because 0.4-0.5 milliseconds isn't very much.)

A Symphony rig with AD16 and DA16 @ 44.1 seem to use 145 samples @ 32 buffer. 145 minus two buffers (145-64) = 81 samples. But there's more: the system. OSX is a multiclient OS, and this takes some overhead. NB: this overhead is included in these 81 samples.
Since the same converters CAN be used for each rig, it makes far more sense to use the same converters to compare performance in this particular case, since it is the native vs. "closed" enviornment types that are being compared, not "pure Apogee vs pure Digi." As such, the converter time in the equation is moot, cancelled out, of no concern, irrelevant...

Quote:
Originally Posted by nativeaudio View Post
Quote:
Originally Posted by nikki-k
Since I established the plugins as each being 1 sample apiece for TDM that I chose, and since the lowest you get with SYmphony is 32 samples, we now know the difference is 30 samples, against the native solution.
It seems that you forget the system overhead: in a native rig, the audio signal travels into the OS and out again. From what I heard a while ago, this does not happen within the buffer size latency. Do we know for sure that there's no overhead baked into those 105 samples PT HD uses either, or if these 105 samples are only AD+DA? Not that long ago, Digidesign sold mBox I/Os meant for the native market, locked to go no lower than 512 samples buffers (times two): that's at least 23 (!) ms latency, not counting the converters. Some TDM users bought that as an addition to their PT rig and concluded that 'native stinks'.
Two points here, and one taken without full understanding of intent...

First, system overhead is minimal IMO. If anything, the native side receives the worst of it, as the TDM system is "closed" and runs within a very strict and rigid manner. "Overhead" is accounted for in "connection" times and such. Typically, these connection times are 1 sample or less at most. It is essentially the same as when you use a hardware digital mixer + hardware digital outboard together. With native, while the OS and such might require some travel time that is outside the buffer, but I would consider that also to be very small in a qualified and properly maintained system.

Second, the original Mbox was created with a certain idea: provide a portable solution that allowed the user to work as if they had an external mixer with them, so as to attain the lowest latency possible while tracking...which would be very close to 0.0 samples, since it is a hardware "pass thru" implementation they used. As such, the drivers were written to allow only a 512 buffer minimum, which is totally acceptable while mixing. Using USB1, this allowed the Mbox to be used for tracking and mixing, and not incur a huge overhead; thus, end users would be less likely to run into problems trying to run at lower buffers. Today, with native power what it is, and USB2, this situation no longer exists...or is not as prevalent...

Quote:
Originally Posted by nativeaudio View Post
Quote:
Originally Posted by nikki-k
If the session/project was large, and had a bunch of plug-ins already, pushing the limits of the system @ 32 sample buffer, then they might have to jump to a 64 sample buffer. Thus, could be double, or more.
True. ANd if a faster Mac comes out tomorrow, allowing for use of a 16 sample buffer setting, the total roundtrip would be lower than today. Will faster Macs be released later this year? Yes. Will Apple release Macs next year that are faster than the 2007 models? Yes.
The buffer is not solely dependent upon the host system: it is also the dev writing the drivers. This was actually a point i saw discussed (albeit very briefly) on the Nuendo forums, and I believe Scott from ADK was one fo the people involved. The point was whether we have hit the point of diminishing returns; CAN drivers and nativ epower reach a point such that it will take an incredible amount of additional native power to attain each tiny step toward the next lowest buffer setting, such as from 32->16? Getting ot 64 was not bad; getting to 32 took a bit more, and working at a 32 buffer is still shaky in most circumstances. What kind of power and genius driver coding will it take ot hit the next tier? Is a 16 sample buffer while wokring even a 20 track session with average plug-in instances realistic with the power we can see coming during the next 3-5 years?

Regardless, we are looking at today, and the very near future, say 1 year. I would wager that we will not see a 16 sample buffer that operates with the systems available by year's close that will allow one to track a full band, or run a decent sized session without severe problems. If at all...

Quote:
Originally Posted by nativeaudio View Post
Quote:
Originally Posted by nikki-k
With no plug-ins being used with the Symphony solution, or many other native solutions, a direct patching scheme can be used, such as ASIO2 provides for, and such as RME has wth their cards/drivers. Thus, only the converters' process time without plug-ins...and PT HD would be the same.
This is correct, if they able to use - and are using - equally fast converters with no added in-between-stuff that takes time, and there's no overhead baked into the mentioned 105 samples from Digidesign. AD+DA on a 192I/0 takes 2,38 ms (if there is no overhead or other time consuming stuff), meaning circa 1.2 ms pr. converter. The AD16/DA16 must use less, because they use eg. 1.6 ms - 154 samples-64 samples (buffers) = 90 samples = 0.93 ms including system overhead. It seems that the best way to get low latency on a PT HD rig is to use Apogee converters.

Right now, what we know this winter, is that the difference for the first time is very small, and slightly lower on a PT HD rig.
Once again, I am using the same converters scenario; since it is feasible, possible and easy to use the same Apogee converters, it is a "cancelled out" part of the equation. The converter process times are both published facts, which no one has been able to provide contrary findings for. So, drawing form these established facts, it is easy to solve the equaton for the unknown variable, being difference.

BTW- this is a point I brought up when the Apogee converters first published the process times: with a native rig using them and obtaining wicked low buffer settings, the difference between that and an HD rig with Digi 192 I/O would be running very close in total trip time latency (possibly/probably less for native!!!), thus the potential recordist would be experiencing nearly the same experience with either solution. Add plug-ins inline though, and the situation slants in favor of the PT HD rig, but only very slightly, and for many (most?), imperceptibly. Thus, with diminishing returns in lowering the buffer size in native systems via more native power and better drivers, so we also get diminishing returns in exactly how great that "lower latency recording with PT HD" benefit is.

Quote:
Originally Posted by nativeaudio View Post
Quote:
Originally Posted by nikki-k
In pure factual terms, yes the 30 sample difference I mentioned is valid...but, once again, it comes down to what YOU are happy with.
Sure, and to clarify: The difference isn't that using one plug-in on a native system adds 30 samples, but that not using no-latency-mode on a native rig (Software Monitoring OFF in Logic) adds some extra latency. Unlike on a PT HD rig, a native rig allows users to bypass the 'audio brain', kind of, meaning even lower latency than we talk about here (but no plug-ins on the recording track).
OK...once again, the point is that making a plug-in active in a native system REQUIRES the audio engine, and thus the buffer, to become a variable in the equation; with no plug-ins, this is removed. As far as the "overhead' part is concerned, I would assert it is low enough to not need to consider. In PT HD, this overhead is a static thing, and is accounted for in "connection times." The PT HD "mixer" is constructed differently than the native one.

Quote:
Originally Posted by nativeaudio View Post
So, only to avoid confusion, adding two plug-ins would NOT mean adding 60 samples. Adding one 'normal' plug-in in a native rig adds zero samples actually, as explained above, and as confirmed by tests (not by midnightsun's test).
I thought I made that clear...maybe not? Inserting ANY plug-in on a track being recorded to "turns off" the direct monitoring feature, however implemented, named, etc. Once a single plug-in of any type is added and made active, that track now incurs the buffer value (at a minimum). The plug-in itself does not add the latency, it is the bypass of the buffer that has been "turned off."

Quote:
Originally Posted by nativeaudio View Post
But as you say, Nikki, the difference between a Symphony rig and a PT HD rig right now is circa 30 samples, or to be more correct: circa 40. (105 samples vs. circa 145 samples). Actually, maybe even 49 samples (1.6ms @ 96k = 154 samples = 49 more than Digidesigns 105 samples, if these numbers are still correct, and also correct @ 96k). I'd like to see more detailed info from both Apogee and Digidesign, not that plus or minus eg. 10 samples by itself matters much in most situations - we're talking about 0,2 milliseconds or less here - it's just my inner nerd that wants some nourishment.
I am sticking with my premise of cancelling out as many variables as possible; in this case, the converters, and doing this by using the same ones for each solution. And, as far as overhead: the PT HD "mixer" is static and an udio track pass thru is sub-sample speed; the native overhead would probably increase (very slightly? alot?) as overall system demand increases. Thus, the benfit PT HD affords in this aspect would also increase...although I think this is minimal and not of much concern these days.

Quote:
Originally Posted by nativeaudio View Post
It's all about which system(s) you want to invest in for the future. I've decided to go the native way, and have friends who have done the same, while others are still on TDM. Good for me, because then I can visit their studios if I need to open one of my hundreds of old TDM sessions to check which plug-in settings I used in my Logic+TDM songs.
Alot of people see the native way as being able to be the best tool for the job; others see PT HD as being the best tool. Sales are up for both, and for someone shopping, it is very good: choices!
Old 4th February 2007
  #104
Lives for gear
 
nikki-k's Avatar
 

Quote:
Originally Posted by nikki-k
Not entirely true...

In any world, certain plug-ins require "X" more time than ANY buffer size due to the type of process it is performing. As an example, a compressor might use a method to look-ahead slightly
Quote:
Originally Posted by nativeaudio View Post
Sure, and that's what I meant by mentioning especially demanding plug-ins (spelled algo in that other post)... :D
There is nothing "demanding" about the process/code; it is a simple matter of the way the effect works.

I had mentioned this prior to your post, which is why it may seem to be the other way around. Besides, I dont know what this "algorithm" thingie is you speak of is...and "algo" is short for it? Is it like Alfie? :D
Old 4th February 2007
  #105
Lives for gear
 
nikki-k's Avatar
 

Quote:
Originally Posted by barryjohns View Post
nativeaudio, I am so glad you joined this topic. You have clarified so much. I appreciate your taking the time to share your knowledge.
I am so glad I waste my time...evidently...
Old 4th February 2007
  #106
Lives for gear
 
midnightsun's Avatar
 

Quote:
Originally Posted by nikki-k View Post
I am so glad I waste my time...evidently...
Quite the contrary, and depends on what you mean by a waste of your time. I have enjoyed reading along and adding a bit. You and nativeaudio have both been generous with your time and efforts to steer this thread toward understanding some of the finer points of comparing Apogee symphomy to PT.

Quote:
Originally Posted by nikki-k View Post
I thought I made that clear...maybe not? Inserting ANY plug-in on a track being recorded to "turns off" the direct monitoring feature, however implemented, named, etc. Once a single plug-in of any type is added and made active, that track now incurs the buffer value (at a minimum). The plug-in itself does not add the latency, it is the bypass of the buffer that has been "turned off."
Though I have actually done some of my own measurements (as I have mentioned eariler) I am a bit cloudy on understanding this critical point you make. Does the addition or insert of each plugin on a given track increase "latency" by system buffer amount additively? It just seems seems that some plugins in Logic Pro add little latency (generic EQ) and others had a tremendous amount of latency (linear EQ). Are there two types of native plugins-- some that stay within the buffer alotment and others that go beyond. I just don't quite have the grasp....

respectuflly,
Old 4th February 2007
  #107
Lives for gear
 
nikki-k's Avatar
 

Quote:
Originally Posted by midnightsun View Post
Though I have actually done some of my own measurements (as I have mentioned eariler) I am a bit cloudy on understanding this critical point you make. Does the addition or insert of each plugin on a given track increase "latency" by system buffer amount additively? It just seems seems that some plugins in Logic Pro add little latency (generic EQ) and others had a tremendous amount of latency (linear EQ). Are there two types of native plugins-- some that stay within the buffer alotment and others that go beyond. I just don't quite have the grasp....

respectuflly,
Yeah, this is a rough one for me to relate.
Many plug-ins do not add any more latency than the buffer already in place adds. Thus, it is not the plug-in or plug-ins adding latency, it is the buffer coming into play when that first plug-in on the track being recorded to is inserted, which "turns off" the direct monitoring ability, which had been effectively bypassing the buffer. Make sense?

Once that first plug-in is inserted, NOW we can contemplate the task the plug-in will attempt to accomplish, and the manner in which it will do so. Some plug-ins will always require more time to process than the time required by the buffer (no matter what the buffer size- it could be a buffer of 1 sample, or 1 million samples). Other plug-ins simply require quite a bt of time to process. This will determine how much latency, or exra trip time, will need to be added to the trip time beyind the converter process and buffer times.
Old 4th February 2007
  #108
Lives for gear
 
nikki-k's Avatar
 

Quote:
Originally Posted by nikki-k
I am so glad I waste my time...evidently...
Quote:
Originally Posted by midnightsun View Post
Quite the contrary, and depends on what you mean by a waste of your time. I have enjoyed reading along and adding a bit. You and nativeaudio have both been generous with your time and efforts to steer this thread toward understanding some of the finer points of comparing Apogee symphomy to PT.
Thx!
I am glad nativeaudio has chimed in, as it has allowed the facts and applications of them to be expressed in a few ways, and has placed light upon several key areas. I feel this allows more people to access this knowledge, since it is explained in different ways, and also provides some debate one might have with themselves when learning.

Having a bad day, sorry...
Old 5th February 2007
  #109
Lives for gear
 
nativeaudio's Avatar
 



Quote:
Originally Posted by nikki-k View Post
Once again, I am using the same converters scenario; since it is feasible, possible and easy to use the same Apogee converters, it is a "cancelled out" part of the equation.
This makes sense.

The reason it's easy to use the Apogee converters on a PT HD rig, is that Apogee has reverse engineered the code needed to make an interface that works with Digi's TDM-gear.

This raises another issue when deciding to go for either a CoreAudio/AudioUnit based system, or a closed system like PT HD. While we now see that performance of native and TDM-systems are very close, one pro for an open system is that one isn't tied up with the hardware or upgrade/price policy of one manufacturer, which would have been the case with PT if Apogee hadn't reverse engineered what they did.

In the native world, if you are unhappy with eg. the performance of MoTu's drivers, or their convertors, you can buy something from RME or Apogee, open all your songs, and they'll be up and running in no time, because they are all based on CoreAudio and Apple's plugin-format. This isn't only a theoretical scenario, lots of people have switched from using hardware from one manufacturer to using hardware from another, because they weren't happy with how their drivers performed. Digidesign was at some point known for making the worst CoreAUdio drivers ever made, (I'm sure this has improved now), and pr. today, their policy is still that one cannot use ProTools with any other hardware than their own (reverse engineering is needed), and you cannot hook up eg. the 192 to other software than PT if you want full functionality.

So when you write that it's it is 'feasible, possible and easy' to use Apogee convertors on a PT HD rig, this is true, but part of the fact behind that truth is that Digidesign is trying hard to make this impossible. If I would have been a new buyer today, I would have taken this into consideration. If RME can make hardware with stable drivers that can be used with all kinds of software, Digidesign can too, but they don't want to. If Digidesign would have supported Apple's plug-in format, AudioUnits, they could have done it,but they don't want to.

Quote:
I thought I made that clear...maybe not?
If you ask me, I for one, think you are totally clear, I was only adding some info in that other post for people who are less familiar with the details than you are.


Quote:
Sales are up for both, and for someone shopping, it is very good: choices!
In the end of 2006, some Digi rep reported that the sales of the TDM-systems went down (14%), and they explained it with the power of Apple's MacPro machines. Actually, I'm relatively sure that they would have offered something 'more pro' (natively based) than the lighter versions of ProTools quite a while ago, if it wasn't for the fact that they are still milking money our of their TDM-gear. They know as much about the power of a MacPro as anyone else, and they also know that there is more to come.

Personally I've promoted use of Digidesigns hardware since the very first version of Pro Tools. Circa 4 years ago it became clear to me that going native would be the way to go for me - but getting there took longer than I anticipated. I blame it on IBM: they weren't able to deliver powerful enough PowerPC chips fast enough. When I finally stopped using Digidesign's hardware, it wasn't because I'm 'against' TDM or something, it was because I felt I didn't need it anymore. I've spent enough money on Digi's various upgrade paths, but most of all, I left the TDM-world for the same reason I stopped using tape (ADAT's): I want to work in ONE audio environment, not two. The hiccups I had with ADAT+Pro Tools came from syncing the two systems up. The hiccups I've had with PT+Logic came from the fact that these systems had to talk together, but wasn't always able to do that well. In most cases, the communication between Logic and DAE went well, but not always. To me (and this is performance-related), most hiccups in a combined system seems to primarily happen between the two systems (eg. native + TDM), not within one of the systems, and this is something to think of because we all want reliable systems.
Old 5th February 2007
  #110
Lives for gear
 
nikki-k's Avatar
 

Nativeaudio- you so hit the nail on the head!
That completely sums up a huge part of the "hate" side of the love/hate relationship for me. While I like the Apogee hardware, I actually do like the Digi 192 I/O alot. I wish they had thought bigger, and made it stand-alone capable, and also thrown a firewire port on it for use with LE. It would allow thos eof us that tire of Digi and PT to move on, and still use the hardware we invested in. Also, it would allow LE users to at least see some sort of worthwhile path to walk when stepping foot on the LE one initially.

The reverse engineering thing- lol- I actually brought this up before when someone asked about what I/O to grab. I think the Apogee stuff is an excellent deal, and would be a great idea (if you like the sound- and es, I think every converter manufacturer has their own "sound" to some extent..until you get up aorund $6-7k per channel of ADC or DAC). But, when considering for PT, there is a chance that when a PT upgrade comes along, you would be sitting on the sidelines for just a little bit while Apogee "breaks" the newest effort of Digi to "keeo it in the family." This was actually my biggest "oh well" when I saw Symphony and all- "Uh...Mac only? Huh." Even though I am using a Mac Pro...for now...which I think is the point...

Personally, I dislike being "forced" to use one thing just to use another. Even though my Les Pauls came with Gibson pickups, it didnt mean I HAD to use those.
Old 5th February 2007
  #111
Lives for gear
 
nikki-k's Avatar
 

Quote:
Originally Posted by nativeaudio
Cured PT addict.
Oh- and love the sig! LOL!!!
Old 6th February 2007
  #112
Lives for gear
 
nativeaudio's Avatar
 

:-)

Quote:
Though I have actually done some of my own measurements (as I have mentioned eariler) I am a bit cloudy on understanding this critical point you make. Does the addition or insert of each plugin on a given track increase "latency" by system buffer amount additively? It just seems seems that some plugins in Logic Pro add little latency (generic EQ) and others had a tremendous amount of latency (linear EQ). Are there two types of native plugins-- some that stay within the buffer alotment and others that go beyond. I just don't quite have the grasp....
Try to think of it this way: some (most) plug-ins are able to do their job within the buffer range (eg 32 or 64 samples), while others are not. Look at this list again:


Quote:
Aphex Aural Exciter 2
Amp Farm 17
Autotune 1
Big Bottom 2
BNR 1537
D-Verb 2
Declicker 8193
Digi Dither 4
DPP-1 2
Drawmer ECL/GCL 6
Drawmer HPF/LPF 2
FilterBank 1
Focusrite D2 1
GRM Resonators 1
Hum removal 2
Lexiverb 2
Lo-fi 405
MDT 4
PAA 1
Pitch Doctor 3925
Pitch Blender 1
Procrastinator 1
Pure Pitch 5116
Recti-fi 2
Sci-fi 2
TC Chorus 1
TC Master X 135
TC Mega Reverb 1
TC Reverb 1
Timeblender 1
Truverb / PAZ 1
PS22 1
Q1 1
Ren comp 65
Ren EQ 1
C1 c/g 341
C1 comp/schain 341
C1 comp 1
De-esser 1
IDR 1
L1 65
MaxxBass 1
All the TDM plug-ins that adds only 1 sample latency would add zero samples latency in a native system. 1 sample = 0.02 ms @ 44.1, so 1 sample = a non-issue.

There's a discussion going on right now - in Digidesigns TDM-forum - started by a PT HD3 user with one Accel card who discovered that he can get about 40 instances of the Waves SSL Channel on his PT setup, while a Mac Pro owner can get 240 instances of the same channel strip plug-in on his native rig. Now that's not a totally fair comparison, because he would also be able to open instances of the same plugin in the (native) RTAS format, but still - with 240 instances without using two HD cards and one Accel card, the question is if he needs more DSP power than the new MacPros can offer.

It could also be said that while he may be able to run 240 instances during mixdown (with an increased buffer setting: buffer settings aren't really important during mixdown), he may not be able to do this when recording, which is also true.
Old 6th February 2007
  #113
Lives for gear
 
tamasdragon's Avatar
 

Also I think it's a question that who needs that plugin count, when tracking something. So all in all, I think this latency thing isn't that big deal. Let's face it, native is catching up. You can get the best emulations from uad, you have fantastic processing power no matter what platform you using (mac/pc).
And don't forget the fact that tdm is a very tight and not the best architecture.
Although a hd is might not double the price of a high quality native, but that few thousand dollars can be spent to very good mics instead imho.

Tamas Dragon
Old 6th February 2007
  #114
Lives for gear
 
midnightsun's Avatar
 

Quote:
Originally Posted by nikki-k View Post
Yeah, this is a rough one for me to relate.
Many plug-ins do not add any more latency than the buffer already in place adds. Thus, it is not the plug-in or plug-ins adding latency, it is the buffer coming into play when that first plug-in on the track being recorded to is inserted, which "turns off" the direct monitoring ability, which had been effectively bypassing the buffer. Make sense?

Once that first plug-in is inserted, NOW we can contemplate the task the plug-in will attempt to accomplish, and the manner in which it will do so. Some plug-ins will always require more time to process than the time required by the buffer (no matter what the buffer size- it could be a buffer of 1 sample, or 1 million samples). Other plug-ins simply require quite a bt of time to process. This will determine how much latency, or exra trip time, will need to be added to the trip time beyind the converter process and buffer times.
This makes total sense. Now when it comes to latency, the lag time as one is monitoring tracking through the software (Logic Pro), for me this is not such a big deal because I am monitoring on the front end with analog. When it comes to track misalignment, i.e. the lag time between and existing track and its bounced print-- this seems to be just a matter of passing through the converters 2 times-- DA then AD and there is little to no software involved. This is an important concept because this is in effect what matters for me when I am overdubbing. Any new track needs to be lined up with the old-- wheter is is being re printed or or just play along to. This misalignment time is what I set for my delay compensation. The tests that I did revealed that I could add all of the plug ins or buffer changes that I wanted and the misalignment didn't seem to change. It would change with sample rate changes. I am going to go through this test again with new insight as brought about in this thread-- my day job is just getting in the way! Any thoughts, disputes, or confirmation of my assumptions. Again I will repeat my tests more methodically when time permits.
Old 6th February 2007
  #115
Lives for gear
 
Tony Shepperd's Avatar
Quote:
Originally Posted by tamasdragon View Post
Also I think it's a question that who needs that plugin count, when tracking something. So all in all, I think this latency thing isn't that big deal. Let's face it, native is catching up. You can get the best emulations from uad, you have fantastic processing power no matter what platform you using (mac/pc).
And don't forget the fact that tdm is a very tight and not the best architecture.
Although a hd is might not double the price of a high quality native, but that few thousand dollars can be spent to very good mics instead imho.

Tamas Dragon

Okay let's look at the statement above and find some answers:
Also I think it's a question that who needs that plugin count, when tracking something. So all in all, I think this latency thing isn't that big deal.
1. If you are doing heavy tracking you might need those plugins. And latency is a big deal.

Let's face it, native is catching up
2. Native IS catching up, it is NOT there yet though.

Although a hd is might not double the price of a high quality native, but that few thousand dollars can be spent to very good mics instead imho.
3. So now that you have bought your good mics, you still have the problems from above when tracking a band, so all the good mics in the world will not help your latency.
Old 6th February 2007
  #116
16942
Guest
Personally for me this is a non-issue as I would never track through plug-ins either way. To me they are for mixing.

Actually the only signal processing I would probably ever track through is a good hardware compressor and maybe EQ if absolutely necessary.
Old 6th February 2007
  #117
Lives for gear
 
nikki-k's Avatar
 

Quote:
Originally Posted by Ashermusic View Post
Personally for me this is a non-issue as I would never track through plug-ins either way. To me they are for mixing.

Actually the only signal processing I would probably ever track through is a good hardware compressor and maybe EQ if absolutely necessary.
I too would not want to track to disk thru a plug-in; however, plug-ins inline or thru an R/S are a definite possibility. As an example, you might have a singer that wants some 'verb. Or, maybe tracking a guitar part that plays on the vibe of a delay being in place. While being able to track with the "sound" present is cool, the flexability we have allows us to track "clean" while using the plug-ins to monitor effected.
Old 6th February 2007
  #118
Lives for gear
 
nikki-k's Avatar
 

Quote:
Originally Posted by nativeaudio View Post
There's a discussion going on right now - in Digidesigns TDM-forum - started by a PT HD3 user with one Accel card who discovered that he can get about 40 instances of the Waves SSL Channel on his PT setup, while a Mac Pro owner can get 240 instances of the same channel strip plug-in on his native rig. Now that's not a totally fair comparison, because he would also be able to open instances of the same plugin in the (native) RTAS format, but still - with 240 instances without using two HD cards and one Accel card, the question is if he needs more DSP power than the new MacPros can offer.
This is the part that I do not like; reducing TDM to a simple matter of "more DSP." (not you nativeaudio, the post in the forum over there) It is more than just that. It is taking something like a digital mixer + an empty effects bay, and building them on the cards; a full enviornment. This enviornment is buildable by the user, and can be built and recalled with each session.

Now, before someone says i am simply praising the Digi gods as a mindless follower...I do not think it is as necessary to have this type of thing anymore. If I had the choice of mixing material I already had written in any studio, I would prefer to be able to use outboard for much of it, if I could magically save all of it in a nice big box. But, since I could not, plug-ins in a digital DAW enviorment are next best. So- since I like the Pro Tools software- would I rather have the PT HD software on a Mac Pro (or similar powerful machine), or would I rather have PT HD software on a machine with an HD7 Accel compliment? Sorry, I would want the latter. And that transcends the simple DSP aspect. 'Cause native is more powerful. For home use? Due to the $$$ involved, I would say gimme the software I love, and not force me to buy hardware I do not need. Love/Hate!
Old 6th February 2007
  #119
16942
Guest
Quote:
Originally Posted by nikki-k View Post
I too would not want to track to disk thru a plug-in; however, plug-ins inline or thru an R/S are a definite possibility. As an example, you might have a singer that wants some 'verb. Or, maybe tracking a guitar part that plays on the vibe of a delay being in place. While being able to track with the "sound" present is cool, the flexability we have allows us to track "clean" while using the plug-ins to monitor effected.
If you can hear siginificant latency on a singer singing through a software reveb on a Mac Pro/Symphony rig in Logic Pro i.e., you have far better ears than me.
Old 6th February 2007
  #120
Lives for gear
 
nativeaudio's Avatar
 

Quote:
Originally Posted by Tony Shepperd View Post
Okay let's look at the statement above and find some answers:
Also I think it's a question that who needs that plugin count, when tracking something. So all in all, I think this latency thing isn't that big deal.
1. If you are doing heavy tracking you might need those plugins. And latency is a big deal.
But adding 'those plugins' does not add latency - a statement which is true for most plug-ins, and equally many plug-ins in the native world as in the TDM world. Nikki is right, entering the mode you need to be in for plugins to be available adds latency (on a native rig), but if Max is right, you'll get only 3ms @ 44.1,while a TDM rig has 2.38 ms @ 44.1. Sound takes 2 ms just to travel from your speaker to your ear, if you sit 2 feet away from your speakers, so 0.6 or even 1.2 ms doesn't make much of a difference. On a TDM rig, you can't leave that 'Software Monitoring' mode, meaning that you can never disable 'plugin mode' entirely, again meaning that you can never get below 2.38 ms @ 44.1kHz.

Quote:
Let's face it, native is catching up
2. Native IS catching up, it is NOT there yet though.
That depends on what you're looking at. Regarding software instruments, the latency of a pure OSX system has traditionally been clearly better than in a TDM system. 3 Symphony cards will give you 192 channels of audio @ 192kHz, a PT HD rig may give you "only"(!) 162 channels. Re. latency, it seems that TDM is still slightly better, but the difference is so small that it isn't noticeable; it's in the same range as the added latency we got when switching from Pro Tools Mix to Pro Tools HD systems. Nobody complained. The one area where TDM never can catch up with native, is price, because extra hardware is always needed, and Digidesign isn't exactly famous for their low prices.

Quote:
Although a hd is might not double the price of a high quality native, but that few thousand dollars can be spent to very good mics instead imho.
3. So now that you have bought your good mics, you still have the problems from above when tracking a band, so all the good mics in the world will not help your latency.
That's true - the only thing is that adding plug-ins doesn't add latency (except in special cases, and in these cases, they add latency on a TDM-rig as well).

It's all about the size of your project. In huge projects, with a lot of tracks, you'll get problems on both a TDM-rig (there are several examples posted in this thread about how few plug-ins a TDM rig can run compared with a native rig on a fast MacPro) and on a native rig. In the example that is being discussed over at the DUC, it was mentioned 240 on a native rig and 40 on an 'extended' HD3 rig: two HD cards + 1 Accel cards. Let's for simplicity say that if all these cards were Accel cards one would get three times as much (you won't) - that's still only half of what a MacPro can deliver natively. Yes, you can buy more cards, and use both native and Accel cards at the same time, and get eg. 480 plug-ins. Brllliant - if you need that much.

Here's the interesting scenario: TDM-users used processing power as a main reason to use PT, and the native users agreed that this was a good point. Then came the MacPros...

Now, many TDM-users used latency as a main reason to use PT - then Apogee popped up with the Symphony-system and recently, it's 'latency-optimized' drivers. Compatibility has also been mentioned, but for many studios, it's more important to be compatible with their non-TDM, Logic using clients, than with other TDM-users.

In the end, it's all about features, workflow, software - even the 0.6 ms latency that may or may not bother someone can be removed by using an analog mixer - after all, that's how engineers dealt with tape head latency during record in pre-DAW times.

Native has more than just been catching up in some areas - eg. Apogee offers higher total track counts @ 192 kHz, but for most people this number isn't interesting at all. Most people will already be able to get the number of tracks they need, a latency they are happy with, the DSP power they are happy with, either from a Symphony or a PT rig, and won't raise an eyebrow if Apple would announce a Mac Pro tomorrow that beats a HD Accel 7 system in all performance tests, because they are happy with what they've got. After years of seeing happy PT-users talking about plug-in and track counts, we'll probably see more posts in the style of that guy over at the DUC who asked "Do you think people buying your records give a rats az whether you used 200 waves or other plug-ins..." a week ago or two.


It does somewhat matter which platform you choose: solution A can offer something you won't get from solution B, and vice versa, but from what I've seen happening the over last 20 years, the differences are becoming smaller and smaller with every update from every manufacturer. Digidesign knows that powerful native solutions are possible, they know what Logic users miss in PT, and Apple know what it is that ProTools has that Logic currently doesn't (Beat Detective and simplicity maybe being the most important 'features'). I doubt that any of these companies have stopped trying to be on top of what's happening. The main difference between Apple and Digidesign is that Apple owns the OS they also develop applications for.

Native/Symphony has been 'catching up' to such a degree that the results now achieved originally were considered 'not possible'. Editor Nick Batzdorf posted this when he saw what a native system now could do, in November: "No way does anything using Core Audio have latency that low," but later posted an apology:

Quote:
I'm mystified about how that could work, given that the converters
themselves are going to take that much time (at 44.1), CoreAudio has
a 64-sample buffer, plus Logic has its own processing buffer that
can't go below 32 samples. But there must be an answer.
That we get what we pay for is only a myth, because what we get for a given price may change from one day to the next, and luckily for Digidesign, a lot of people will not trust that the Symphony system can deliver what they promise. A main reaction when Apogee announced Symphony was 'can they really deliver this?' - Nick Batzdorf wasn't the only one questioning if this is possible. The situation (native replacing TDM) is somewhat similar to the reactions Digidesign got when they once announced that PT/hard disc recording is 'pro' enough to replace tape machines.

I think that's why Digidesign is hesitant about delivering a 100% native version of ProTools (with all it's features): if Digidesign can deliver this, then the diehard TDM fans will 'get' that the competition can deliver a totally professional, native solution as well. Digidesign's main income has been selling DSP cards, not software, and maybe they simply can't afford loosing that bit yet. There may be more people using native PT versions (than using TDM-versions) right now, but Digi aren't making as much money on these rigs, because they don't come with DSP cards costing $4,000 each. If you're a salesman, it's more fun to sell something costing $12,000 than a selling software costing $1,000 only.

Digidesign will catch up with Logic re. a few things, and Logic will catch up with PT. regarding a few things too. Apogee has more than 'catched up' with the sound quality of the 192 I/O a long time ago, and Digidesign will certainly compete more on the native market than they do today, I don't think they can afford not to. It's all about when.
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