|
|
Post by audiospecific on Dec 22, 2023 20:39:49 GMT -6
The buffer shouldn't matter here though. I'm not sure why it's even quoted. DSP RTL isn't even entering the computer. Unless I'm not understanding what buffer you're referring to? Yeah I thought it was weird that Avid was quoting it too. Maybe it’s because Avid’s figures are from within PT, while UA or Apogee’s figures are using their own standalone monitor mixers. I didn’t think HDX armed-track monitoring would have to stream through the DAW buffer, but them including it in their quoted latency suggests that maybe it does. This would suggest that putting the buffer at something reasonable (like 512) would add another 5ms of latency (at 96kHz). Can that be true?
The whole thing is what made protools pro tools isn't incorporated in their hardware.
The hardware mixer generations (I'm talking about tdm core with 888/24 24 mix systems) is they were PCIe digital mixer cards with add on DSP cards that communicated between each other on the PCIe back plane.
So you only had conversion and sample buffers for latency. so samples at the hardware clock level is 44 samples for 44.1Khz 48 samples for 48 Khz
For instance:
hardware loopback of 888/24 I/O A/D/A = 79 samples Instigated by "monitor button"
When used as a "hardware insert" this only added 2 samples to this for a total of 81 samples a route to and from a dsp bus using a plugin 6 samples a route through dsp mixing bus is 5 samples
So what they are selling is just a fancy overpriced m-box.
And I know there are studios still running those cards on version 8 and don't care. But they have real outboard too.
|
|
|
|
Post by kcatthedog on Dec 22, 2023 20:58:04 GMT -6
I don’t believe I said it was not needed in all cases. i said if you are are not tracking with UA plugs, and only mixing with them , then the native versions are attractive. Don’t need apollo or cards so get better bang for you bucks as you don’t need an interface with dsp or a satellite. So, why buy into the UA dsp works now if you don’t have too? Depends on your workflow. Fair enough. It just sounded like a blanket statement about DSP when you said "The only reason to buy UA dsp now is to run the uad2 plugs which are not native: yet". That sounded pretty absolute to me. Well, that is the only technical reason: uad2 needs dsp , Uadx doesn’t. For me, just mixing with plugs, UA dsp is irrelevant. I understand the latency concern: just not a problem I have, as a singer songwriter. not pretty, but easy enough to turn off all plug ins or bounce a stereo sub mix to a new session, record new tracks and export that back to original, there are workarounds . |
|
|
|
Post by ragan on Dec 22, 2023 22:19:06 GMT -6
Yeah I thought it was weird that Avid was quoting it too. Maybe it’s because Avid’s figures are from within PT, while UA or Apogee’s figures are using their own standalone monitor mixers. I didn’t think HDX armed-track monitoring would have to stream through the DAW buffer, but them including it in their quoted latency suggests that maybe it does. This would suggest that putting the buffer at something reasonable (like 512) would add another 5ms of latency (at 96kHz). Can that be true? No, running the buffer at 64 or 1024 during tracking, has no effect on HDX monitoring latency. That’s what I would expect. Seems odd they feel the need to qualify the latency figures with test condition DAW buffer. |
|
|
|
Post by Quint on Dec 22, 2023 22:32:02 GMT -6
Fair enough. It just sounded like a blanket statement about DSP when you said "The only reason to buy UA dsp now is to run the uad2 plugs which are not native: yet". That sounded pretty absolute to me. Well, that is the only technical reason: uad2 needs dsp , Uadx doesn’t. For me, just mixing with plugs, UA dsp is irrelevant. I understand the latency concern: just not a problem I have, as a singer songwriter. not pretty, but easy enough to turn off all plug ins or bounce a stereo sub mix to a new session, record new tracks and export that back to original, there are workarounds . Ok. It didn't sound like you were just talking about yourself, but you've now clarified that you were just talking about yourself, as it pertains to DSP. Got it. |
|
ericn
Temp
Balance Engineer
Posts: 16,098 
|
Post by ericn on Dec 22, 2023 22:33:53 GMT -6
No, running the buffer at 64 or 1024 during tracking, has no effect on HDX monitoring latency. That’s what I would expect. Seems odd they feel the need to qualify the latency figures with test condition DAW buffer. I don’t think thy meant to qualify as they meant to show the conditions that the spec was for. |
|
|
|
Post by Quint on Dec 22, 2023 22:37:39 GMT -6
That’s what I would expect. Seems odd they feel the need to qualify the latency figures with test condition DAW buffer. I don’t think thy meant to qualify as they meant to show the conditions that the spec was for. They might as well tell us what color of shirt the person was wearing when they did the latency test on the DSP too. That is if Avid is interested in giving people irrelevant information.  |
|
|
|
Post by ragan on Dec 22, 2023 22:40:03 GMT -6
That’s what I would expect. Seems odd they feel the need to qualify the latency figures with test condition DAW buffer. I don’t think thy meant to qualify as they meant to show the conditions that the spec was for. If it has no effect on the figure they’re advertising, it’s a weird thing to include. |
|
|
|
Post by audiospecific on Dec 22, 2023 23:53:53 GMT -6
No, running the buffer at 64 or 1024 during tracking, has no effect on HDX monitoring latency. That’s what I would expect. Seems odd they feel the need to qualify the latency figures with test condition DAW buffer.
Depends on where the monitoring takes place. Either at the interface DSP section of the chip or in software. That is why the monitor button in the DAW doesn't work well in most systems. Because they are looping it in and out of the computer. Something like total mix would be at the interface chip dsp But something doing that like a UA box with DSP plug on top should have only a couple sample buffers as latency instead of media latency looping back and forth via the usb/TB port with a software controller environment. Which something tells me that is what they are doing.
|
|
|
|
Post by drumsound on Dec 23, 2023 0:19:29 GMT -6
That Avid figure is at 64 sample buffer. I’d be interested to know how many people are actually running at 64 sample buffer. In my experience, not many. I think the meaningful difference between HDX and Apollo is workflow (if you’re using PT, HDX wins there) and I would hazard a guess that people who think they’re hearing latency difference between Apollo and HDX are actually not. But, that’s a guess. I was talking to a guy who does regional live gigs, he was telling me that since going digital wireless and IEM their are a bunch of acts who always complain about latency, so much so they have been buying up used analog just to cut it down. I know that Lyle Lovett had (has?) a no digital consoles clause in his rider. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 23, 2023 10:18:23 GMT -6
That Avid figure is at 64 sample buffer. I’d be interested to know how many people are actually running at 64 sample buffer. In my experience, not many. I think the meaningful difference between HDX and Apollo is workflow (if you’re using PT, HDX wins there) and I would hazard a guess that people who think they’re hearing latency difference between Apollo and HDX are actually not. But, that’s a guess. most of the people who say they need HDX are using some godforsaken old Mac or Windows DAW desktop, that if updated or running a plug from after Bush years would keel over with a 64 sample buffer. There are not that many post GW Bush HDX plugs anyway.. The buffer shouldn't matter here though. I'm not sure why it's even quoted. DSP RTL isn't even entering the computer. Unless I'm not understanding what buffer you're referring to? Yeah I thought it was weird that Avid was quoting it too. Maybe it’s because Avid’s figures are from within PT, while UA or Apogee’s figures are using their own standalone monitor mixers. I didn’t think HDX armed-track monitoring would have to stream through the DAW buffer, but them including it in their quoted latency suggests that maybe it does. This would suggest that putting the buffer at something reasonable (like 512) would add another 5ms of latency (at 96kHz). Can that be true? The current HDX hybrid engine admits it is a glorified Apollo for monitoring mixes. HDX uses TI DSP chips vs UAD’s Analog Devices SHARCs. Yes, raising the buffer greatly increases the latency. The interfaces, converters, and digital consoles that claim ultra low latency than is typical for the connection protocol used usually compromise the anti-imaging and anti-alias filters to achieve that with iir filters of varying quality, short fir filters with inadequate band rejection followed by an iirc filter, or some on chip fir filter with a horrible transition band. The lowest latency interfaces are still RME and Lynx PCI-e cards but their internal conversion is horrible and subject to all kinds of noisy garbage from the computer. Hooking up some real converters to them adds latency but still one of the cheapest low latency 8in 8 out setups is a used Lynx AES card with used Yamaha AD8HR for the pres and adc and a good, half decent hifi converter (Chi-fi best avoided) or two for monitoring. An entire setup like that can have higher latency than a good thunderbolt interface but is much cheaper than a couple Apollos or a tricked out Lynx Aurora or Apogee Symphony. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 23, 2023 10:34:53 GMT -6
The whole DSP and apollo vs hdx is an apt comparison and somewhat amusing... Amusing in that for many, many years Avid was basically a closed system and now, as of a few years ago with the whole Luna/apollo combination... UA basically created yet another closed DSP system. And ya know what? I'm glad they did! Its nice to have a choice. And if you can't make a record on either rig you can't make a record period. Me? I made another choice and don't run either... but I do have a UAD satellite and genuinely adore it. Would buy it again tomorrow. It does add horsepower to the system. Full stop. Not even a question. Love my virtual outboard rack! I think its probably right to speculate about the future pricing of DSP plugs... rightly so. But also right in that as a DSP user? With my rig I don't buy a whole lot. Its been a year since I added a plug? Always think I should take advantage of the end of year sale... But its kinda dumb to think they would completely abandon that entire DSP ecosystem and go 100% native. Not at this point in the process. I also think its equally dumb to think there aren't ways UA as a company could separate me from my money. How many things do they make now besides DAW software?! Guitar pedals. Microphones. That ox powersoak DI thing... Man I'd love to have a pair of those. Anyone got $3k I could borrow? Need a record mixed?!! Its a big company now. Not just a handful of guys building 1176 & LA2A reissues in a little rented shop anymore. I have a pair of those... from those early days... I also have new UA gear. Saul good! They're good I'm good the gears good what's not good? Besides paying for all that awesomeness? Any competition for Avid is a good thing, but to compete with Avid you really need to be in the post world. I keep wondering if SSL’s latest masters might take a stab at it. Imagine a system where for video you record on Sound Devices edit and mix in a SSL / Harrison intergrated DAW. They could compete with AVID Eucon with their own protocol. the new ssl controllers are better built than their current analog equipment but still use the old Mackie protocol. Their current plugs aren’t great either. They’re often worse than 20-25 year old Waves and Sony algorithms. Only now does SSL have any channel strips that can compete with UAD2 but 10 years too late, there are much better native plugs that aren’t emulations, and the emulation guys have moved onto modeling weirdo unique behaving and distorting stuff you cannot get the analog of like Plugin Alliance, Shattered Glass, Fuse, Plugin Alliance, Goodhertz, even UAD versus something relatively commonplace that is still made in some bastardized form |
|
ericn
Temp
Balance Engineer
Posts: 16,098
|
Post by ericn on Dec 23, 2023 10:39:57 GMT -6
I don’t think thy meant to qualify as they meant to show the conditions that the spec was for. They might as well tell us what color of shirt the person was wearing when they did the latency test on the DSP too. That is if Avid is interested in giving people irrelevant information. In the world of broadcast / post it is still pretty standard that if you state a figure you state the relevant variables in testing that the figure was mesured under. To this day it irks me that manufacturers publish and advertise an amps wattage to the MI guys never stating conditions or available current. Sure that amp will deliver 2000 watts, but read the fine print at 1K at 2 ohms one channel driven, oh it’s going to pop most breakers because it is going to draw 35 amps. |
|
