Page 4 of 7 FirstFirst 1234567 LastLast
Results 31 to 40 of 64
Like Tree70Likes

Thread: I Challenge UA to Fix The DSP Limitations on This Platform

  1. #31
    Heavy UAD User caustix's Avatar
    Join Date
    Sep 2015
    Location
    N.Ireland, UK
    Posts
    760
    Mentioned
    0 Post(s)
    Quote Originally Posted by shimel View Post
    Dan, I'm 100% agree : I'm myself an quiet happy Apollo Quad + Satellite octo's user ( despite the fact I have to use native plugins because my 12 DSPs are a not enough power )
    My posts here are just to understand and guess where UA is going, and what are their technical options for the near future
    I explicitly told about some entry level Apollo's new owners because I know people that are upset when the realize what are the real limitations of the UAD plateform
    I have 12 too, happy UAD user and content. But I think I'm beginning to agree with the case for more power or cheaper satellites - it's not a deal breaker at the moment. I just do this for a hobby.

  2. #32
    Moderator DanButsu's Avatar
    Join Date
    Jul 2012
    Location
    Montreal, Canada
    Posts
    4,843
    Mentioned
    11 Post(s)
    Quote Originally Posted by shimel View Post
    Dan, I'm 100% agree : I'm myself an quiet happy Apollo Quad + Satellite octo's user ( despite the fact I have to use native plugins because my 12 DSPs are a not enough power )
    My posts here are just to understand and guess where UA is going, and what are their technical options for the near future
    I explicitly told about some entry level Apollo's new owners because I know people that are upset when the realize what are the real limitations of the UAD plateform
    Oh I agree as well, I have the same setup, 12-chips, and would like 8 more! All this talk about better chips, new chips, native... maybe UA could streamline the code while keeping Distressor quality emulations? Then we'd get more for our money.

    I'd love to setup an SSL Channel Strip on every channel, but you can only fit 2 per chip, ok for a jazz session, but not much else!
    shimel and caustix like this.
    Apollo QUAD Silverface TB | OCTO/Magma TB | UAD v9.4 | Mac mini 2011 i7 Quad Sever | OS X 10.8.5 | Pro Tools 12.5.2 | Studio One 3.5 Pro | Harrison Mixbus 32C v4 | Amphion one18 driven by a Vecteur A6 HiFi Amp | Adam Sub10 | Subpac M2x | AlphaTrack | 32-Channel Soundcraft Ghost | Sony APR5003v 1/4" two-track Reel to Reel | Roland RE-200 Space Echo | DBX 120X-DS Subharmonic Synthesizer | Sample of my work: dansoundsmith - Soundcloud

  3. #33
    Experienced UAD User
    Join Date
    Dec 2015
    Location
    Los Angeles
    Posts
    108
    Mentioned
    2 Post(s)
    Quote Originally Posted by scratch17 View Post

    The maximum DSP available per channel is one chip . This is an amazingly restrictive limitation.

    Here is a real world example of how drastic the affect of these limits becomes. You have an Apollo 8 Quad. You are recording an acoustic guitar part in mono. Nothing else is being recorded in the session.

    You want to run it through a channel strip to get the Unison microphone preamp and do some EQ. Maybe you need a bit of compression, too.

    The acoustic guitar you are using still sounds a bit thin. Better add Woodworks in the next slot.

    And of course you want to place the guitar in the correct spot in a great room. I have a small one room studio so I always slot the Ocean Way Studios plugin on my acoustic guitars. I also like the integrated reverb that matches the room's space.

    So I put the SSL E Series Channel Strip in the Unison slot. My DSP goes to 41.8%. Then Sound Machine Woodworks goes into the next slot. DSP is now at 78.2%. When I try to add Ocean Way Studios I get a warning that I am unable to instantiate it because I have exceeded DSP load limits.

    In the meantime, I am sitting with three DSP chips at idle. With the restrictions of one chip per channel, the additional chips I paid for when I bought my Quad are about as useless as tits on a bull. That is just plain wasteful. Note that this example does not even account for the DSP usage required to run the system.
    I assume this is through UAD console, is it the same on the mixing end? Say, protools channels?

    Also, is there a basic chart/primer somewhere on the DSP/chip/sharing limitations beyond the single chip per channel issue? And when using multiple DSP hardwares (i.e, multiple satellites and/or multiple Apollos?)...Or are there none?---besides instances limitations?

  4. #34
    Experienced UAD User
    Join Date
    Jul 2015
    Location
    Arlington, Texas
    Posts
    308
    Mentioned
    2 Post(s)
    @Gerk said:

    SHARCs are not the problem here, it's a serial vs parallel issue.
    Gerk, you make great points. But I am not sure that they are insurmountable without changing the laws of physics. I will be thinking outside of the box here. If my thinking goes awry because I missed some fundamental fact (or got one completely wrong out of ignorance), please forgive me in advance.

    To sum up, I started this thread with two major issues I have with DSP allocation on Apollo.

    1. One chip per channel, regardless of how many are available.
    2. No DSP sharing from one Apollo to a channel on another.

    Having pondered Gerk's post, I have changed my mind about issue 2. I now believe it to be impractical.

    Gerk may well be absolutely correct. There may be zero technology currently available to fix issue 1. If not, a simple statement from UA that they are working on a solution would go a long way from my point of view.

    For me, the inability to use more than one chip on a channel is almost a deal breaker. I want this fixed, and I believe it is possible to do so.

    I understand the input / output requirements of both processing methods. I went back to school in the 2000's and got 4.0 in C++ and Java. While this was primarily high level software programming, I had to take assembly (low level machine language). It was the one course in the curriculum that scared me. I got through with help from some young classmates who were far better coders than I.

    The assembly class was a beginner's course in machine language. We did some serial machine language coding. Parallel processing was above that class' level. But we went over the basics.

    So I believe I understand both concepts, at least on a fundamental level.

    Gerk, I do not put myself at your level of understanding in this area, especially the hardware architecture aspects. And I have no DSP programming experience. So I may easily be missing some basic issue.

    Emulating the inserts on analog consoles is potentially part of the concept I have for a solution. Bear with me as I digress and get really basic for a bit. Relevance will become obvious.

    Everyone here should already know that an analog mixer is a parallel set of serial analog data streams, routed by default to an analog summing amp. Because the data is analog, no parallel processing overhead is required. In fact humans cannot hear latency from input to output on analog mixers. I am just using this as an illustration of the base concept that designers of digital audio interfaces must emulate.

    A digital audio interface with A/D converters must have inputs for analog data streams that are laid out in parallel with one another. The requirements of serial digital data streams in parallel with one another does add processing overhead from input to output. And of course, the overhead added by A/D and D/A conversion is a factor that adds latency and truly is a matter of the laws of physics. Thus the round trip monitoring latency of around 2 ms in modern interfaces.

    Whether it's an analog mixer or a digital audio interface, the serial streams might have data modifiers such as filters, compressors, limiters, etc. Without re-routing the stream's path, those modifiers cannot be used on any of the other serial streams. The situation is akin to the UA DSP limitation of one chip per channel.

    Let's look at the serial path in a digital audio interface when data modification is the responsibility of on board DSP. In an Apollo 8 Quad, any one of the four chips is "assigned" to any one of the 8 input channels by the Apollo's operating system. "Assigned" actually means that the data stream is re-routed so that it passes serially through the DSP chip, and then is routed back to the original stream's path. Think of the assignation process as if it were an insert point on an analog mixer. The DSP chip is simply inserted into that loop.

    This routing change through the DSP chip must add some overhead (processing cycles, re-clocking so that other streams are in sync, etc.) and thus some small amount of latency. But the stream stays serial even though the data is digital. Considering that overall round trip latency through an Apollo is close to or under 2 ms, the added latency by rerouting the stream through a DSP chip can't be a huge factor.

    So why not make it possible to assign two DSP chips in serial to a single stream? This would be like adding two insert points to a channel, one after the other.

    While a single plugin's load likely could not be shared across the two chips, at least you could slot more power hungry plugins onto a single channel. And with two chips per channel maybe 6 DSP slots per channel makes more sense. If you wanted to use more than four plugins on a channel, the two added slots would make sense.

    Based upon other threads here, the problem lies in PFGA limitations. It is not a parallel processing issue.

    It may be that the PFGA's input / output matrix is maxed out. If the PFGA limitations in current Apollos is indeed the real issue, a hardware change will be required to increase the size of the matrix.

    But what if UA designed an Apollo that had the same size matrix, with fewer inputs and outputs. This would free up input and output points in the routing matrix for digital insert points which could add the flexibility required to put DSPs in serial on a single channel.

    So why not create an Apollo that is limited to 1 input per two DSP chips routed in series by default onto each channel? With that architecture, the current PFGA chips UA utilizes might be adequate. This would at least solve the issue of one chip per channel.

    Note that this would cause no obsolescence of any currently available hardware in the UA lineup. Yes, Console would have to be re-written to add the new routing capabilities of the new model. But nothing else would be affected.

    Optional insert point position could also solve the issue of certain limitations on plugin order.

    For example, others here have asked for the ability to insert a Townsend Sphere plugin before a Unison channel strip. That is not a possibility with the current requirement to have Unison before other insert slots. An insert before the Unison hardware would let you load the Sphere plugin before a Unison channel strip.

    Now if you noticed, I suggested a default two chips per channel. What if you wanted more DSP to do parallel effects processing with a mono input to a stereo output? As long as the PFGA matrix points were available, this might also be possible.

    How about side chain insert slots? Anyone here ask for that?

    If UA were to deliver a quad 2 in / 2 out Apollo with the ability to do 2 chips per channel, I'd buy it.
    Apollo 8 Duo (BF), Apollo Quad SF, iMac i7 Quad, Logic Pro X, Studio One 3 Pro

    Gibson SG, Hamer Duotone, Brian Moore i213, Taylor 710 BCE
    Mesa Recto Pre + 20/20, 68 Bandmaster

    Mackie Onyx 1640, Event SP-8, Kemper Profiling Amp, Roland VG-99

  5. #35
    Experienced UAD User
    Join Date
    Jul 2015
    Location
    Arlington, Texas
    Posts
    308
    Mentioned
    2 Post(s)
    @ohzone asked:

    I assume this is through UAD console, is it the same on the mixing end?
    If you load a plugin into Console's slots, all of the restrictions are in place, whether you are mixing or tracking.

    When you load a UA plugin into a DAW insert slot, the Apollo operating system assigns the DSP load to the available hardware using UA's specific algorithm. I think I read somewhere in one of the manuals that the algorithm does the most efficient job possible of load allocation. That does not mean the algorithm is unlikely to have some restrictions on which DSP gets what plugin load.

    Note that if you want to use a Unison plugin, it must be loaded into Console, not a DAW slot.

    I have never seen the specifics as to how the algorithm does its job. Anyone here have any clue?
    Apollo 8 Duo (BF), Apollo Quad SF, iMac i7 Quad, Logic Pro X, Studio One 3 Pro

    Gibson SG, Hamer Duotone, Brian Moore i213, Taylor 710 BCE
    Mesa Recto Pre + 20/20, 68 Bandmaster

    Mackie Onyx 1640, Event SP-8, Kemper Profiling Amp, Roland VG-99

  6. #36
    Experienced UAD User
    Join Date
    Dec 2015
    Location
    Los Angeles
    Posts
    108
    Mentioned
    2 Post(s)
    Quote Originally Posted by scratch17 View Post
    @ohzone asked:



    If you load a plugin into Console's slots, all of the restrictions are in place, whether you are mixing or tracking.

    When you load a UA plugin into a DAW insert slot, the Apollo operating system assigns the DSP load to the available hardware using UA's specific algorithm. I think I read somewhere in one of the manuals that the algorithm does the most efficient job possible of load allocation. That does not mean the algorithm is unlikely to have some restrictions on which DSP gets what plugin load.
    Not sure if I understood the response here----are UAD plugs, inserted into a DAW (say protools), also limited to a max of 1 DSP chip per channel?

  7. #37
    Moderator DanButsu's Avatar
    Join Date
    Jul 2012
    Location
    Montreal, Canada
    Posts
    4,843
    Mentioned
    11 Post(s)
    ^ ohzone, no they are not!
    Apollo QUAD Silverface TB | OCTO/Magma TB | UAD v9.4 | Mac mini 2011 i7 Quad Sever | OS X 10.8.5 | Pro Tools 12.5.2 | Studio One 3.5 Pro | Harrison Mixbus 32C v4 | Amphion one18 driven by a Vecteur A6 HiFi Amp | Adam Sub10 | Subpac M2x | AlphaTrack | 32-Channel Soundcraft Ghost | Sony APR5003v 1/4" two-track Reel to Reel | Roland RE-200 Space Echo | DBX 120X-DS Subharmonic Synthesizer | Sample of my work: dansoundsmith - Soundcloud

  8. #38
    Experienced UAD User
    Join Date
    Jul 2015
    Location
    Arlington, Texas
    Posts
    308
    Mentioned
    2 Post(s)
    As I said above there is a proprietary algorithm that decides on DSP load allocation when a UA plugin is loaded into a DAW (like Protools). I do not know and cannot say whether the limitation wrt one chip per channel is operative in this case. But I suspect that it is also true when you load a UA plugin in a DAW. This is based on supposition that the limitation comes from the UA hardware, specifically the PFGA, not the Console app.
    Apollo 8 Duo (BF), Apollo Quad SF, iMac i7 Quad, Logic Pro X, Studio One 3 Pro

    Gibson SG, Hamer Duotone, Brian Moore i213, Taylor 710 BCE
    Mesa Recto Pre + 20/20, 68 Bandmaster

    Mackie Onyx 1640, Event SP-8, Kemper Profiling Amp, Roland VG-99

  9. #39
    Practically a UAD Expert Kcatthedog's Avatar
    Join Date
    Dec 2011
    Posts
    8,279
    Mentioned
    13 Post(s)
    "For me, the inability to use more than one chip on a channel is almost a deal breaker. I want this fixed, and I believe it is possible to do so."

    Ah, its not broken and ah why ? how many plugs ins does one need to use per channel when tracking anyways?
    slamthecrank likes this.

  10. #40
    Experienced UAD User
    Join Date
    Jul 2015
    Location
    Arlington, Texas
    Posts
    308
    Mentioned
    2 Post(s)
    Dan, thank you for disabusing me of my incorrect supposition!
    Apollo 8 Duo (BF), Apollo Quad SF, iMac i7 Quad, Logic Pro X, Studio One 3 Pro

    Gibson SG, Hamer Duotone, Brian Moore i213, Taylor 710 BCE
    Mesa Recto Pre + 20/20, 68 Bandmaster

    Mackie Onyx 1640, Event SP-8, Kemper Profiling Amp, Roland VG-99

Page 4 of 7 FirstFirst 1234567 LastLast

Similar Threads

  1. Marshall plugin DSP limitations and work-around
    By ChadAustin in forum General Discussion
    Replies: 10
    Last Post: 19th January 2017, 11:44 PM
  2. Replies: 5
    Last Post: 16th January 2017, 11:08 PM
  3. UAD2 Cubase Problem - How to Fix? Plugins Do Not Lower DSP When Turned Off
    By D2CStudios in forum Support & Troubleshooting
    Replies: 3
    Last Post: 2nd June 2013, 12:36 PM
  4. Softube, please start a DSP platform
    By djsynchro in forum Unrest Department
    Replies: 5
    Last Post: 22nd March 2010, 12:33 AM
  5. Softube, please start a DSP platform
    By djsynchro in forum General Discussion
    Replies: 0
    Last Post: 31st December 1969, 06:00 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •