Yes...the H9000 is real!

Discussion in 'The Rack Space' started by hendrik7, Dec 19, 2016.


  1. chlorinemist

    chlorinemist Member

    Messages:
    403
    Joined:
    Jul 20, 2016
    the heck? I'm not telling you my assumptions. I'm simply describing the facts of the matter, based on well-known facts, as well as a combination of both their statements in these press releases and conversations between myself and Eventide. I have said nothing worth provoking such a rude and arrogant retort. But alas

    I have actually asked people at Eventide working on the H9000 the question you were asking: will the H9000 do bigger algos than the H8000? The answer was a simple NO. Their explanation confirmed what I had come to understand about Eventide for a long time, and was along the lines of the answer I gave you when you asked the same question.

    Correct me if I'm wrong: Eventide's business model has been focused almost entirely on repackaging their existing algos into new formats and reinventing their approach to user interfacess for over ten years now. I have been paying attention, and I imagine you have as well, and it seems to be clearly a fact that at least the vast majority of their products since H8000 are repackaged versions of algorithms from the era of the H8000 or earlier. Often simplified versions. Eclipse, Factor pedals, H9...

    The question about monolithic modes is still up to debate, i don't know the answers there but I'm not expecting some huge library of brand new algorithms that do brand new effects. If that's supposed to be a surprise, they've been doing a hell of a job keeping the R&D a secret. Dunno why they'd hide such valuable marketing information though. And to go so far as to lie and tell me they aren't currently focusing on bigger algorithms? Why/how would they keep the secret after doing these press releases, giving the release date, selling a limited amount of H9Ks and all? Not seeing it, personally.

    I can agree with you on this point
     
    Last edited: Oct 15, 2017
    rsm and bosetuno like this.
  2. italo de angelis

    italo de angelis Member

    Messages:
    4,509
    Joined:
    Mar 21, 2011
    Well..

    thank you for reporting what THEY say. It wasn't clear that was their words with some dressing of your opinions.
    At the moment it looks they haven't got it right. Improve the UI? Yes. Improve the machine possibilities? Nope. That's wrong.

    I'll correct you on the Eventide "biz model"... yes you are completely wrong.
    Eventide has always been a strong and unique forge of ideas materialized into amazing stuff in their products. Any and every new machine had a wave of new things added to what had been done before.
    People using these processors for decades do expect this kind of consistent innovation in these science toys... not just using a ****ing iPhone do to the same things as before.
    The H8000 introduced a TON of new things to the previous line of products: 5.1 effects, stereo to 5.1 effects, MIDI Virtual Racks, Multiple Machines managed as MVRacks, the new H8000 reverbs, Custom Scales pitch shifter with graphical interface (music staff) and a lot of other things, dispersed thru existing banks. There was never a repackage based product. Most pre_existing algorithms have always been improved/enlarged when more power and new modules have been available, from the 4000 to the 7000/Orville platform and over the H8000. This has always been a main concept in advancing.
    And the same is thru for Eclipse. Advancing thru different OS versions, including new algorithms not available on the H8000. Vintage Delays was created on Eclipse and made it sell a LOT... so they had to port it over the pedals.
    The repackaging is for the pedals! Oh yes... there little development and a ton of cannibalizing from the good rack treasure. Once in a while a "new" effect... mostly existing stuff with a different name.
    The amount of energy and work about creating new modules and new algorithms with them was just amazing back in the days. You have no idea how many cool things were made... and worked... and never saw the light.

    All Eventide advertising has always based on keeping the heritage in showing the new things possible on a new product, from the 3000 to the 4000, to the 7000 to Orville, on to the H7600/H8000.
    When the DSP4000 was released it was based on about 90 modules (Vsig). By the time the platform reached the H8000 status... those modules got to 200, introducing modules you could only dream before, like Reverb A which could finally use any number of delay lines, from 4 to 32 (in group od 4) and addres up to 32,5 secs of delay memory to create density, or all delays and shifters modules which could finally address from 10 to 30+ seconds of delay memory instead of 640 ms which forced you to string a ton of them and add math modules to do a stupid knob to control them all... wasting tons of DSP rsources and Vsig lines (Vsig has a limitation to the max number of modules you use in an algorithm... it's not endless), or double precision filters (that's a LOT of power being taken and the H9000 won't be able to use them as it should), or the fantastic Harmonix module...PitchTime which does independent Time and Pitch compression/expansion, cool comprtessors to do multiband compression for mastering work, the Distortion Preamp for any kind of audio mangling, based off multiple user selectable curves, the retriggerable oscillators and the H3000 lfo, the dozens of new interface modules to create a control no other unit offers to the user.... dozens of modules, progress, improvement, BIG ONE... with plenty of DSP engineers doing actual code work.
    All a DSP engineer HAS to do... IS creating modules. That's THEIR job!
    So now we have a SHITLOAD of power... running the same old code and modules and the existing presets? Why? Because they concentrate on the UI? Gimme a ****ing break, dude. 13 years have passed doin' what? Counting beans and peanuts!
    I really hope FOR YOU that they have been missing the critical info on their press releases so far and the good stuff is in there or you have no idea how sorry you'll be. As it is now... it doesn't look like a substantial change/improvement over past products for a creative head like it's always been. If you move on to a 7 strings guitar with divided pickup... you'll be stuck as you wouldn't believe today.
     
    Last edited: Oct 15, 2017
    rsm likes this.
  3. italo de angelis

    italo de angelis Member

    Messages:
    4,509
    Joined:
    Mar 21, 2011
    MISSING MODULES
    many of them have been requested for ages... by paying customers

    gimme:
    -a true 5.1 and 7.1 matrix based reverb... for real surround work
    -a surround panner as you cannot do it, yet!
    -a multitap delay the designer can choose the number of delay lines (2 to 64), with modulation on each line, hi/low cut filters in each feedback, an internal mixer with levels and pans.
    Add feedback breakout connections for each delay so that anything external can be inserted there. This module alone can create entire products!
    -consolodate filters in single modules for graphic and parameteric EQs so that one doesn't have to go crazy stringing single bands in an algorithm and exhaust resources pretty quickly.
    -a modern LOOP module that can do time stretching, can address any amount of on board delay memory.
    -also make a stereo LOOP module like above.
    -add musical control of "non in key" notes in the Diatonic PitchShifter!!!
    -finally build a MIDI Note controlled pitch shifter that plays the right chords with a keyboard!
    -build a COMPOSER module that uses the sampler to record a set time of audio... you can divide in pieces, cut, store, edit, reverse, pitch shift, time stretch... and reassemble them as you like in a new "form"... how cool that would be???
    -the samplers and LOOP modules? They are in desperate need of being able to store samples, recall/reload them, edit them... what's the point of such improved UI then?
    -implement real sample rate mayhem based effects... if you want to really get "vintage"! Make the hardware ignore digital I/O as a choice... to use these old_new tools!
    -gimme a ****ing distortion curves graphic editor so that I can design crazy stuff there, store it and use in the distortion preamp.

    Just as an appetizer....
     
    rsm, Man with Gas and bosetuno like this.
  4. rsm

    rsm Member

    Messages:
    13,810
    Joined:
    Jul 7, 2008
    Location:
    On an island in the Atlantic Ocean
    To summarize my oversimplified understanding to see if I got this right:

    So far it appears the H9000 is providing additional parallel processing (horizontal scaling) for recording/playing multiple tracks and unlike the H8000 (and previous) it no longer provided combined processing (vertical scaling) for creation and use of larger, more complex algorithms?

    If I understand this correctly, then how difficult would it have been to provide vertical scaling as the previous products have had while also providing the increased horizontal scaling, even if both were limited or not possible simultaneously? Seems it would cover what both sound creators/designers and studio engineers want/need/use?

    What am I missing?
     
  5. Gone Fission

    Gone Fission Member

    Messages:
    246
    Joined:
    Feb 24, 2006
    Speculatively:

    Yoking together separate processors to work together as one sometimes isn't an easy trick. The Orville didn't manage this--only the H-8000 added this capability even after updates , which suggests that support hardware is needed to pull it off. This is Eventide's first go of doing a multi-process ARM machine. Maybe on update if buyers are lucky. Or on the follow-up machine if not.
     
  6. italo de angelis

    italo de angelis Member

    Messages:
    4,509
    Joined:
    Mar 21, 2011
    Well.... this is getting into a bit of headache as we don't have ALL the official info, so everything is mostly based on speculations. I'm sure we'll get to know what the facts are in the near future.

    -FACT : the H9000 has 4 quad cores processors
    -FACT : the H9000 can run 16 algorithms/presets

    Speculation: the unit has 16 cores (4 chips x 4 cores each)- It would be pretty obvious an algorithm/preset will always run on a core.
    What does that mean? It looks like a hardware imposed bound. Any structure can be as large as a single core allows it to be. How large is the question?
    Well...Eventide says those algorithms are "H8000 class" ones, which means they are as big as an H8000 allows them to be.
    So far one would simply think the H9000 is just a collection of 8 H8000s in a box, running their 16 presets.

    But... there are other FACTS to consider:
    -the real "H8000 class" meaning
    -the impact of sample rate on algorithm size

    "H8000 class" is a bad definition as it's incomplete. The H8000 has a lot of preset running on a single DSP (the box has 2 DSPs). Thuse you can run two of them at the same time.
    A monolithic mode was implemented; it "blends the 2 DSPs into a single giant one to make basically one of two things possible:
    -run a single preset at higher sample rates (88.2/96KHz). A lot of those presets running on a single DSP need the extra power of monolithic mode to run at higher sample rates. You loose number of presets running at thesame time and you gain higher quality audio processing... when needed. So you find a lot of presets in two versions, normal and monolithic.
    -run a really giant algorithm that would not run on a single DSP, no matter the used sample rate. That's just the only way you can make something really big... work... and sometimes even up to the higher sample rates.
    It appears then that the H8000 has a dynamic way to allocate processing power and resources to accomodate algorithms size and/or used sample rate. That's a pretty smart and useful way to use power, I believe.
    Why? Because I can finally build bigger structures than what I could do on Orville and make many of the Orville larger presets run at higher sample rates. So we are getting more crativity at work (vertical) too.

    So... what doeas really mean "H8000 class" algorithm for the H9000? We don't know... but I have a feeling the definition is about the regular ONE DSP running ones.
    If that's the case... we are loosing "power per algorithm" instead of gaining it, as in every new product taking the place of a previous one. I can't run the big monolithic algorithms that would exceed the single DSP power at lower sample rates AND higher sample rates. That's a bad limitation.
    It may well be possible that the H8000 can run a larger algorithm than the H9000, at the same sample rate, in monolithic mode... a bit embarassing to my way of looking at these matters. Only one algorithm, I know, but LARGE!
    One other important aspect comes into the picture here... how about sample rates? It's expected the H9000 will be able to run 192KHz sample rates (if not higher ones... some of the optional I/Os formats can run higher than 192KHz rates).
    We don't have ANY onformation about that. What happens to an algorithm size when 192KHz is used? Not that this would really mater to any and every user... because very few ones would really need to do that. Even 96KHz is not the usual way you will run your box... BUT... when needed/required by a project... you gotta know what the consequences are... BEFORE YOU BUY!

    More speculations...
    -are those cores capable to go monolithic? In simpler words... you could be able to load MUCH bigger algorithms even though you loose in quantity.
    I would be VERY happy to be able to load "only" 8 algorithms (that's still a shitload of tremendous power!) thanks to cores getting monolithic in couples. This would allow some vertical improvement, not too much, over the H8000.
    If those cores could get monolithic in 4s... like a whole 4 cores chip runs ONE single algorithm... that's where I would expect the real vertical gain over previous units.
    Considering the limitations higher sample rates impose on real power... that would make the difference. BIG one.
    But we have no information at all about monolithic mode(s) on this unit. My feeling is that it doesn't do monolithic... as it's a quite complex thing for such architecture.

    It's probably the case that the main thinking in developing the H9000 was the studio and the software developed to intergate a powerful processing platform into the DAW workflow.
    The idea could also be that existing algorithms are already quite large that should cover most production needs.
    But if that's the case... those are lesser than smart assumptions because on such open platform you are allowed to build things or modify existing ones.
    What happend if you need to add something to an existing algorithm, a large one, that already takes the whole single core power'? You can't.
    The solution is not to string another preset to the previous because you may need something "nested" into the effects structure of the first... not before or after it.
    That scenario becomes hell!

    So... until we get all the official info about "H8000 class algorithms" definition, sample rates limitations on their size, monolithic yes or no... there is not much we can be sure about.
    Different scenarios are possible... and some of the above speculations can be very real IF those limitations work in those ways.

    I can see studios, most of them, being super happy with the kind of power and size of the H9000.
    But if those limitations are real... I can see many potential customers getting H8000s instead, particularly among musicians.
    They will get cheaper and one may end up having 2 or 3, or even more... to really have the real vertical power to get algorithms made as they need, rather than to be able to load 16 algorithms that would never work in the same way.

    We have to wait...
     
    Last edited: Oct 22, 2017
  7. italo de angelis

    italo de angelis Member

    Messages:
    4,509
    Joined:
    Mar 21, 2011

    The 4000 did that (monolithic) but you didn't notice it . The Orville was made with separate DSPs but some people wanted monolithic in those days... that's why the H8000 was done the right way. But don't forget the Orville single DSP was running a preset 4 times larger (and faster) than a 4000... so even without monolithic, you still had both HUGE vertical improvement and horizontal too, as two presets were now running at the same time.
    I'm afraid this is a much harder thing to do on multicore architectures... but would really love to be proven wrong.
     
  8. markiv2290

    markiv2290 Member

    Messages:
    22
    Joined:
    Jan 12, 2015
    Location:
    Mirabel, Canada
    Agreed 100%.
    The new modern architecture, GUI, available fx chaining are things a creative musician want. I actually sent a suggestion to Eventide on making a "light" version of the H9000. Even half the processing power would probably be plenty.
     
    italo de angelis likes this.
  9. italo de angelis

    italo de angelis Member

    Messages:
    4,509
    Joined:
    Mar 21, 2011

    Definitely MORE than enough power... 8 algorithms on a single box? One could kill for that.
    The thing is... if they make a 50% down scaled model and price it 4500$ vs the big one at 6000$... it doesn't make sense to buy the small one... as they always wrongly did in the past with the DSP7xxx and H7600. The difference in processing power must be reflected in a more logical price difference. The small brother should not be more than 3500/3800 $.
     
  10. markiv2290

    markiv2290 Member

    Messages:
    22
    Joined:
    Jan 12, 2015
    Location:
    Mirabel, Canada
    Agreed!

    Further, I could see how they could scale this trough the product line: 2 algos version, 4 algos version, 8 algos version and the full throttle all-out 16 algos version, with a pricing strategy that make sense with this power scaling (understood that referring to "algos" as a measurement unit is over-simplifiying things a bit, but you get the idea).

    All with the same 2U chassis, GUI and modern platform.
    There's some economy of scale to be made on the production line, R&D / product development synergy naturally happening, all the same time meeting the needs of various customers (guitarists vs studios etc.) having various DSP power requirements and the accompanying budget limitations :D

    I do hope Eventide are hearing my prayer :)
     
  11. italo de angelis

    italo de angelis Member

    Messages:
    4,509
    Joined:
    Mar 21, 2011
    They won't do this.
    Eventide is a very small company. They don't have the power to do this kind of scaled work.
    Maybe, maybe a 50% power cut version may happen, but I wouldn't loose my sleep on it.
    Unfortunately!
    We'll see....
     
  12. Gone Fission

    Gone Fission Member

    Messages:
    246
    Joined:
    Feb 24, 2006
    I'm guessing supply chain/parts obsolescence for the H8000 is part of what made the H9000. (Although I might underestimate the hi def surround user base.) The same end-of-life issues will catch up with the Eclipse eventually. That and convergence on ARM processors will probably make the Eclipse replacement a chip off the H9000. I'm guessing it will be very similar to an Eclipse--two algorithms, some additional routing flexibility borrowed from the 9000, but probably fixed algorithms rather than VSig support (or whatever the next edition is called).
     
  13. kingbobalou

    kingbobalou Member

    Messages:
    199
    Joined:
    Jun 17, 2015
    Oh boy here we go with the fukery BS.
     
  14. italo de angelis

    italo de angelis Member

    Messages:
    4,509
    Joined:
    Mar 21, 2011
    @kingbobalou
    not really fukery BS 'cuz the person posting before you *knows* what he's talking about very well.
    That's very different from the 90% of action going on on other subforums here, truly based on "fukery BS" from people who have no technical idea when opening their mouths.

    @Gone Fission
    hard to know what they are going to do. Replacing Eclipse should be a daughter_derivation of the whole work made for the 9000. That simplifies things a lot.
    The ARM processor choice? I still need to hear that because I'm not sure about its virtues. What I know is that it's the cheapest chip out there... and that's NOT giving me any thrill in the bones.
    They might just decide to build a single line of machines, all based on 1, 2 and 4 ARMs (that's 4, 8 or 16 cores) and call the single chip version "H900" or more likely something else to not create confusion.
    And probably lock algorithms editing down on it, making it a 4 fx engines and be done with it... no IT I/Os, scaled down UI (smaller display, etc.).
    What would be interesting is if they made the 4 cores working as 2 fx engines, so... like on Eclipse you can load 2 algortihms. Which would mean BIG ones, like double than the H8000... and if that's not possible on the H9000...
    it'll be all about pissing customers off!
     
    Last edited: Jan 12, 2018
    ctreitzell likes this.
  15. pangea2003

    pangea2003 Member

    Messages:
    114
    Joined:
    Apr 21, 2015
  16. Allthesound

    Allthesound Member

    Messages:
    525
    Joined:
    Apr 19, 2012

    Really interested to get more info on this, I'm saving to buy a Harmonizer now (if i can ever stop buying PCM's) was pretty much settled on a used H8000 or Orville.... But this! ^^ According to Eventide a beta of a Vsig editor for this will be ready around the same time the first units ship. If i go for it it will probably be the unit with the blank front panel. They said they hope to be able to offer the remote app on mobile devices as well as the PC/Mac versions. If i could control it from my iPad that would be huge! NAMM should answer alot of questions hopefully.
     
    ctreitzell likes this.
  17. italo de angelis

    italo de angelis Member

    Messages:
    4,509
    Joined:
    Mar 21, 2011
    NEVER NEVER NEVER get blank units! In time you may regret it very badly and lose money.
    A machine like this is not something you buy lightly and keep for a couple of seasons. This is keeping stuff... in time you won't have a way to use it because any remote software can't be supported in years on IT devices. In 10 yrs from now... you won't have any remote.
    People still uses 30+ yrs old units with their dedicated hardware remotes OR display/knobs (not blank units). Think about this very carefully as most don't get the point.

    Also... never buy something on promises. Companies can promise things that are never accomplished.
     
    ctreitzell likes this.
  18. Allthesound

    Allthesound Member

    Messages:
    525
    Joined:
    Apr 19, 2012
    I agree with you about buying something on a company's promises , i have been burned before and am about to get fooled again if i can help it.
    As far as whether or not buying the blank panel version is a good idea depends on a few things. granted ideally its best to have both front panel and remote control,
    but with a price difference of $2k between the two it's not trivial. I have inquired with Eventide to see if they plan on offering the front panel controls as a upgrade to the blank units, waiting to hear back now.

    For the way i plan to use it the front panel would get used very little if at all. I'm not worried to much about being able to run a remote app 10 or even 20 years from now on a PC there is always a way to run legacy software if you really need to.
    The H8000 is 13 years old now and they still plan to provide updates and bug fixes to Vsig for it even after H9000 launches , i don't doubt Eventides long term commitment to the H9000 platform.

    For me by the time any of this becomes and issue i will be either to old to care or worm food.
     
    ctreitzell and chlorinemist like this.
  19. chlorinemist

    chlorinemist Member

    Messages:
    403
    Joined:
    Jul 20, 2016

    From what I've heard from Eventide, iPad/iPhone control is 100% for sure going to happen. They've already implemented this in the H9 pedals and they mentioned that tablet control was the main reason behind their decision to go ahead with making an H9000 blank faced model, since they think many people will prefer using tablets in general since they have a much larger display. I think I will be going to with the blank panel myself. I find it very hard to believe Eventide would just eventually brick these devices rather than figuring out some permanent control solution if the iOS solution proves no longer usable for some reason (unlikely imo...).

    At this point I think there's no question the H9000 is the way to go over the previous models. I have an H8K and it's maximum of 2 simultaneous algorithms is a pretty huge limitation at the end of the day, especially compared to your typical DAW workflow or modern processors like the (gasp!) AxeFX series, which can create chains of dozens of effects easily.
    The H9K ups the limit to SIXTEEN simultaneous algos. Completely new ballgame. Honestly I believe it will make its predecessors look downright primitive in comparison, which is a massive feat.
     
    ctreitzell likes this.
  20. italo de angelis

    italo de angelis Member

    Messages:
    4,509
    Joined:
    Mar 21, 2011

    H8000 is 15 yrs old. An no... they won't provide any bug fixes as there is no need to. You won't see any OS updates.
    Of course Vsig, if it will happen, will have its course... the question remains if it will thw new version will support the older platform, as the modules have been ported to a completely different platform. There is no knowledge if one will be able to load any sigfile (from previous unints) and Vsig will understand them and order the H9000 the right compilation. We'll see... a lot of talk so far and little facts...
     
    ctreitzell likes this.

Share This Page