• Register
  • Help
Page 1 of 3 123 LastLast
Results 1 to 10 of 28

Topic: Speed-boosting the technology...

  1. #1

    Speed-boosting the technology...

    Here\'s a thought..

    We all know about all the big libraries coming our way, and there are many big libraries available already. I\'ve always been in favor of small but useful libraries as opposed to bloated less useful ones because I didn\'t have 20 giga pcs to load them up in. Now I realize that I owe people such as the ones involved in the Vienna project, and QLs newly announced Symph. Orch. library, a BIIIG thanks. I think these kind of projects are gonna help immensely in pushing technology one step further.

    Never before have I witnessed an askewed kind of development boost like the one we\'re facing. I think it is truely remarkable that mere library developers are actually writing history and forcing the software programmers (of which we are 100% dependent on) into breaking new grounds. I tell you now, Tascam needs to come up with a whole new product, hardware requirements have to be met by the industry (harddrive architecture has to be completely re-thought and re-designed), and everything needs to be affordable.

    Let us not forget how everything started. The great thing about Gigastudio was that you no longer needed a rack of 10 E4\'s to make orchestral music. You kept everything inside one or two PCs and that was it.
    Soon the library developers were getting weird ideas in their heads. \"I wonder how a dedicated 1gb instrument would sound!\" - the technology was there, and thus it was done. (slightly overlooking pianos as I\'m sure you understand)
    Lately we have seen an explosion in the market, filling our hds to the rim with more or less useful collections of samples, while the actual gigastudio technology has not improved a bit.

    We\'re in desperate need of THREE serious improvements in Giga 3.0:

    1) Polyphony needs to be at least tripled.
    2) The amount of midi-ports needs to be AT LEAST doubled.
    3) Support for uneven bits (like instruments compromising 3,5,7 etc. layers)

    *1 - This will require a whole new harddisk technology, and preferably a cheap one as we\'d also need a humongous amount of space. CPU is not really an issue. We have

    *2 - It\'s not that hard to implement an 8-port interface really. One that can run seamlessly over a normal 10/100 TCP/IP network and also function with traditional (although terribly outdated) midi interfaces.

    *3 - This would be to effectively save polyphony and create instruments with less unecessary programming limitations.

    So Tascam, get your a** in gear and get started on the next generation of samplers, pronto!


  2. #2

    Re: Speed-boosting the technology...

    Agreed. Wonder how much difference the serial ATA idea will make?

    Hope these new libraries will incorporate lite options for those of us who are hardware poor!

  3. #3

    Re: Speed-boosting the technology...

    Also, as long as GS is limited to 1GB RAM, these libraries will be way too huge to really use unless you split them up over many GS machines.

  4. #4

    Re: Speed-boosting the technology...

    I haven\'t made a thorough investigation of the other sampler platforms out there. Is there currently anything that significantly rivals GS in performance for utilizing these large libraries? If so, perhaps the high end library market might need to migrate to a different platform.

    Lee Blaske
    <font size=\"2\" face=\"Verdana, Arial\">I would like to hear what the Unity Session users have to say about this.


  5. #5

    Re: Speed-boosting the technology...

    I get the impression from folks around town that orchestral emulation is still a relatively small, niche maket. More people use samplers, like Halion and EXS24, to make electronic music. Smooth integration into the sequencer is what many people like about those products.

    I think it is asking a lot of Tascam to invest significant research dollars to develop a product that serves a smaller market and niche developers.

    I\'m just curious...does Giga read data off of Firewire drives? If it doesn\'t, that would be the most significant thing they could do, in my opinion.

    That way, you could buy these new libraries pre-loaded on FW drives with DVD or CD back-ups. Also increasing the memory would be handy.

    However, I must say that I almost never use more than 40 midi channels at once as I transfer to audio before my channels are all filled up. I run out of memory far quicker than I run out of midi-channels, especially with GOS.

    I\'ll likely buy a second PC for another Giga shortly. That seems more managable to me than working with one \"super-giga\" machine.

    A 2ND or 3RD PC seems to be the way to work with these new Super Libraries. Spread them out over different PC\'s connected to a drive network.

    That way, if one crashes and burns, you can still meet a deadline, at least. With just one PC, you are really flying without a net.

  6. #6

    Re: Speed-boosting the technology...

    \"I get the impression from folks around town that orchestral emulation is still a relatively small, niche maket. More people use samplers, like Halion and EXS24, to make electronic music. Smooth integration into the sequencer is what many people like about those products.\"

    From what little I can see, I get a similiar impression, though it\'s a relative thing. It seems to me that non-orchestrally centered folks prefer the Halions, but if you are someone seriously considering creating orchestral mockups, Giga seems the more popular choice. And what market has Tascam been hitting up? Well, its been a while, but the last AD I remember seeing for Giga had Hans Zimmer\'s face slapped on it.

    Orchestrally-focused people may be a small niche in the world overall, but I think its a niche Giga has been really focused on, and unless they become suddenly disinterested in what I\'d imagine is a large chunk of their user base, I could imagine them changing alot for it, as is being suggested here.

  7. #7

    Re: Speed-boosting the technology...

    Money being the bottom line as always, of course tascam would cater to the full spectrum, but really, the changes being addressed here wouldn\'t only affect orchestrally minded folks.

    Features such as integrated networking capabilities for midi data transfer would save everyone hundreds on their setups, even thousands depending on how many machines they\'re running. I\'ve been using Midi Replicator for this, but it\'s not a perfect system, and there are some sync problems I\'ve run into now and again. It needs to be better. Midi cables are archaic, but no one seems to be pushing this technology on a large scale.

    Polyphony, as Thomas mentioned, is a huge deal for most everyone I know. Modern processors, ATA drives, cheap DDR ram, running on 533mzh FSB machines are capable of delivering more in alot of cases, but still we hit the software limitation of 160 voices, and then it\'s time to add a whole other machine with license. When this software was first released, I can understand garanteeing users a certain response time from key-stroke to sound, as well as preventing a system from getting bogged down with too much real-time work and being deemed unusable/unstable by its user base, and perhaps this was a tool to help secure that. But things have rapidly gone beyond that now. Surely the poly count max could be doubled at least, if not tripled, to allow for those users that have newer systems capable of utilizing it. As we can already hit the poly max with only 4 ports (and do all too frequently), I also agree it should be raised as well. Windows supports 10 midi ports open at a time I think?

    Basically, streamline the software so that\'s its adaptive to newer technology. Right now I\'m running a system that can run as though it were at least 1.5 giga machines. Just the other day I ran Nuendo with a 3 hour unedited digital cut of a short film I\'m helping with, 17 tracks of 24bit audio spanning 18 minutes each with realtime effects on all of them, and gigastudio was hovering around the 100 (plus) line with regards to poly, every now and then peaking out (hence I had to mix some of those tracks down to the aforementioned 17.) I don\'t want to run all of this on one machine, but my second comp\'s down at the moment until the new soundcard arrives. The point is, I don\'t have anywhere near the best equipment available, but if it can handle all of this, I have no doubt it could handle 2 gigastudio\'s worth of polyphony right now. But the software doesn\'t allow for this.

    This isn\'t just orchestral stuff, this is the kind of thing everyone needs. Better integration with more headroom. (of course... if I suddenly found myself with giga software that wouldn\'t peak out until it hit a 480 poly count, I\'d probably start looking for faster hardrives and motherboards that would support them with more overall bandwidth. [img]images/icons/rolleyes.gif[/img]

    Along that firewire hard drive question, what about USB 2? Doesn\'t it have like 866mb/sec transfer rate or something nuts like that? Most new mo/bo\'s have at least 4-6 USB 2.0 ports on the back. I\'d kinda like to know if GS handles either, since I can\'t physically shove any more HD\'s into this machine.

    I suppose the bottom line in my mind is simply this - Physical technology will continue to grow exponentially following the release of 3.0. The software should allow for that. (it\'s a great time to be alive!!!)

  8. #8

    Re: Speed-boosting the technology...

    The thing with HArd drives, is that its not really the bus transfer rates, its the access times.

    Most firewire drives are still running at ATA speeds. I ahven\'t seen one that spins at 10,000 RPM.

    Firewire and USB are awesome in terms of transfer rates, but they are only that. Transsfer rates.

    This would only really make a difference if you had multiple hard drives all with redundant data (like RAID) but Giga could recognize this and stream data of them sequentially/serially (like chadwick stated)

    thats a lot of HD space for a single instrument [img]images/icons/smile.gif[/img] You might as well buy more giga machines.

    TJ is right it will require a new Hard Disk altogether, not jsut taking advantage of bus speeds.

    Or it will take reqriting code.....or us dealing with horrible latency....

    Which is an option to be honest. How many instruments do you really play in \"real time\"? ONE! (not including layers) If Giga could sync up to a sequencer through some sort of timing loop, it could use a larger buffer length/read ahead for samples that are being played back via sequncer, and prioritize for the one real time instrument to be played. This might take some weight off of the acess time issue. And could even span out to offering DX/VST effect support for \"non realtime\" tracks.

    yah...I\'m dreaming

    another thing about hard drives. People say computers have been advancing and Giga hasn\'t, but the truth is, only processers really have. Hard Drives have essentially been the same for quite some time. Since Giga relies so heavily on the hard drive, its easy to see why things ahven\'t advanced at the same speed as the processers.

  9. #9
    Senior Member Bruce A. Richardson's Avatar
    Join Date
    Sep 1999
    Dallas, Texas

    Re: Speed-boosting the technology...

    OK, guys, mind if I step in for a little reality check here?

    1) Any person can have 20 ports of GigaStudio technology running right now, for less than the cost of a mediocre used car. Using cheaply available networking technology, such a system will function as a seamless whole. People are doing it all over the world right now. Works like a charm.

    2) DAW software is quite capable of streaming 100+ tracks of audio in and of itself. Not everything has to be \"virtual.\"

    3) Isn\'t the message here really, \"I want more capability, but I don\'t want to pay for it?\"

    There isn\'t a single capability needed to run any library today or tomorrow that isn\'t completely addressed right now. All you have to do is add machines. That isn\'t Giga\'s problem. It\'s not Tascam\'s problem. It\'s not Maxtor\'s, Western Digital\'s, or Microsoft\'s problem. That\'s your problem baby.

    You don\'t see any of your \"heroes\" in here whining about polyphony or ports, or memory, or sample producers, do you? No.

    Food for thought.

    Love and kisses,

  10. #10

    Re: Speed-boosting the technology...

    Bruce, so what you\'re saying is basically \"screw technology advancement and development - we got what we need already.\" ?

    Sorry, but I think that\'s a pretty narrow-sighted perspective on things.

    What lan software is capable of running 20 midi ports simultaneously? I\'m only aware of Midi Replicator, which is limited to 8, but soon to be expanded into 16 ports.

    Sure DAW systems are capable of running 100+ tracks but I thought Simon explained the harsh reality to you already, in some other thread. DAW software like Logic and Protools etc. cannot stream 100+ tracks with 3ms latency if the audio tracks are not stored in an aligned order on the harddrive. All this boils down to seektime. If you are playing 64 different .gig\'s at the same time, the harddrive will have to physically seek back and forth between clusters to read to buffer and while the harddrive cache helps somewhat, the effective seektime you\'re looking at is somewhere around 8/9ms*64 (in a worst-case scenario) which adds up to 510-580ms. In reality the average seektime would be somewhere around 11ms (over 700ms) per buffer. Giga uses ram buffers to overcome this problem to some extent, but if Giga 3.0 uses the same engine as 2.5 it would be unrealistic to expect more than 200 voices, MAX, unless you increase the ram buffer (and kind of defeating the whole purpose of disk-streaming.

    Face it, hardware technology isn\'t quite ready for this yet.

    \"3) Isn\'t the message here really, \"I want more capability, but I don\'t want to pay for it?\"\"
    This is what you have to expect in any kind of industry. It\'s all about constantly improving a product, and keep costs down. Competition is the keyword here.

    \"You don\'t see any of your \"heroes\" in here whining about polyphony or ports, or memory, or sample producers, do you? No.\"
    I should think they would rather contact Tascam directly with their issues at mind rather than blurting them out on a public forum. Even if I could afford 20 giga pcs, I would never buy them. It\'s hot enough in my room already, running 4 pcs and a bunch of studio equipment with the window open 24/7. Air condition is not an option. Keeping the noisefloor down is of utmost importance and a noisy working enviroment is not tolerable. Not to mention the cable-spaghetti I\'d be facing with a 20 pc setup.

    Bottomline: Akai\'s used to be the industry standard. Then EMU with their E4\'s came along. Now Gigastudio captured most of the market share with orchestral producers.
    It\'s called technology advancement, and it is constantly evolving.

    If TASCAM doesn\'t improve their product, some other developers will see the opportunity. Our need for improvement is insatiable. In the end the customers always win.

    Think about it.


Go Back to forum

Tags for this Thread


Posting Permissions

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