Topic: Hauptwerk and Organteq comparison?

I have a friend who uses Hauptwerk organ software.   How might this be better than Organteq.   I am already a user of Pianotech Studio
Ian

Re: Hauptwerk and Organteq comparison?

in a nutshell

Hauptwerk offers wet sampling (stereo/surround) of an instrument and its acoustics. It is a sampling-based development platform. There are many independent producers and the quality is very variable, not always satisfying! Currently the sampling library is very large. If the recordings are done well then it is very realistic. In short, the easiest way to capture the reality of that particular place.
Dry sampling is reduced to a few useful cases and has long since been abandoned because it is too complicated to achieve compared to market interest. The semi-dry sampling (with tail truncation) is not as satisfying as a true dry sampling and you probably get a better or comparable result with Organteq without the reverb.

If you are interested in a dry sound then Organteq currently has no sophisticated multi-channel routing management like Hauptwerk. This is important for distributing the allocation of sounds and greatly reducing the intermodulation distortions that result from the digital mix of dry sounds.

In perspective with Organteq you can shape the sound from within with a level of clarity that can be difficult to achieve with microphones. Try to think what you can do with Pianoteq! (Making the parallel with the piano I prefer the Garritan CFX but the level of clarity of Pianoteq is superior)

Re: Hauptwerk and Organteq comparison?

Hauptwerk is from a technical perspective very different from Organteq. The former is based on samples from real organs while Organteq is an algorithmic  simulation. No recorded audio there, but software modeled to behave like an organ. Both are good in different ways. I don’t have enough experience of both nor a real organ to compare, but Organteq is the winner for me because I believe in physical modeling technique and I see sampling as an outdated method.

If you are not interested in the technical aspects I think you just need to download demos of both to play with and hear which one is best for your taste.

Re: Hauptwerk and Organteq comparison?

Many thanks to you both for your replies.   My friend unlike myself is a very accomplished professional pipe organist and academic and  has played numerous European and USA organs.   I have sent him a link to the Organteq webpage and he will soon listen to the demos.   It was him who mentioned Hauptwerk to me.
Ian

Re: Hauptwerk and Organteq comparison?

Beemer wrote:

I have a friend who uses Hauptwerk organ software.   How might this be better than Organteq.   I am already a user of Pianotech Studio
Ian


Probably, there is a different use case scenario for each application. With HW/Grand Orgue/jOrgan you already can choose from many different organ types and sounds, OrganTeq has yet to develop or give users tools to develop the many different types people would want to play. In theory, OrganTeq could be the one and all application for building VPOs once it has matured due to, in theory, infinite possibilities to emulate pipes.

I wouldn't say at this stage that HW has become obsolete for the simple fact that it offers plenty organs to choose from to suit every taste and budget.

Re: Hauptwerk and Organteq comparison?

I'm a user of Hauptwerk (Advanced version) and Pianoteq 6.  Having now tried the free trial version of Organteq, one major issue for me (amongst many, I'm afraid to say) is maximum polyphony.  As a test, I tried playing Mulet's Tu Es Petra with all stops out (and manuals coupled).  Within the first bar the very unpleasant "distortion" happened due to audio / polyphony overload with the default setting.  I was using a MacBook Pro quad-core i7 from 2016 using built-in audio.  I had to reduce the polyphony to 96! to completely avoid any overload.  A question for Modartt would be "what system / processor is required to handle polyphony of 512 notes?"

To give you another comparison, one of the restrictions of Hauptwerk BASIC edition (v4) is a maximum polyphony of 1024 notes whereas the Advanced edition is unlimited.

Re: Hauptwerk and Organteq comparison?

harnham wrote:

I'm a user of Hauptwerk (Advanced version) and Pianoteq 6.  Having now tried the free trial version of Organteq, one major issue for me (amongst many, I'm afraid to say) is maximum polyphony.  As a test, I tried playing Mulet's Tu Es Petra with all stops out (and manuals coupled).  Within the first bar the very unpleasant "distortion" happened due to audio / polyphony overload with the default setting.  I was using a MacBook Pro quad-core i7 from 2016 using built-in audio.  I had to reduce the polyphony to 96! to completely avoid any overload.  A question for Modartt would be "what system / processor is required to handle polyphony of 512 notes?"

To give you another comparison, one of the restrictions of Hauptwerk BASIC edition (v4) is a maximum polyphony of 1024 notes whereas the Advanced edition is unlimited.

It's normal, it happens to me too with a mac mini quad core 2011. I suggest you decrease "Sample rate" to 48kHz and increase "Audio buffer size" to 2048. If this is insufficient then you need to take a recent desktop PC with many many cores. We await the indications of Modartt.

The maximum polyphony of 512 is relatively high because with Organteq the sounds are dry (reverberation is an effect) and do not have a tail like with Hauptwerk (with wet sampling) so the comparison must be done in this way. Polyphony consumes much more CPU with Organteq than with Hauptwerk and you will probably saturate the CPU first!

Re: Hauptwerk and Organteq comparison?

aile wrote:
harnham wrote:

I'm a user of Hauptwerk (Advanced version) and Pianoteq 6.  Having now tried the free trial version of Organteq, one major issue for me (amongst many, I'm afraid to say) is maximum polyphony.  As a test, I tried playing Mulet's Tu Es Petra with all stops out (and manuals coupled).  Within the first bar the very unpleasant "distortion" happened due to audio / polyphony overload with the default setting.  I was using a MacBook Pro quad-core i7 from 2016 using built-in audio.  I had to reduce the polyphony to 96! to completely avoid any overload.  A question for Modartt would be "what system / processor is required to handle polyphony of 512 notes?"

To give you another comparison, one of the restrictions of Hauptwerk BASIC edition (v4) is a maximum polyphony of 1024 notes whereas the Advanced edition is unlimited.

It's normal, it happens to me too with a mac mini quad core 2011. I suggest you decrease "Sample rate" to 48kHz and increase "Audio buffer size" to 2048. If this is insufficient then you need to take a recent desktop PC with many many cores. We await the indications of Modartt.

The maximum polyphony of 512 is relatively high because with Organteq the sounds are dry (reverberation is an effect) and do not have a tail like with Hauptwerk (with wet sampling) so the comparison must be done in this way. Polyphony consumes much more CPU with Organteq than with Hauptwerk and you will probably saturate the CPU first!

Same observation here. Tutti with all couplers engaged is impossible, but then my music PC is quite old and underpowered. For now, I can play tutti not coupling anything just fine. Probably future versions could make use of CPU power more efficiently? Down the road it should be an i7 or i9 I suppose.

Re: Hauptwerk and Organteq comparison?

Tutti wrote:

Same observation here. Tutti with all couplers engaged is impossible, but then my music PC is quite old and underpowered. For now, I can play tutti not coupling anything just fine. Probably future versions could make use of CPU power more efficiently? Down the road it should be an i7 or i9 I suppose.

I've noticed that duplicate instances of stops stack, like the Trompette 8' which is part of the Pedal, Grand Organ, and Recit divisionals.  On a regular organ, the ranks for this stop would only be in one divisional--likely the Grand Organ, while having optional redundant stop knobs on the other manuals, which means that when a note is played on another of the manuals or via a coupler, the organ still will only make the sound once.

In this version of Organteq, the sound is simulated multiple times through different tuning settings and layered, which both makes the sound a little louder--which wouldn't happen with a real organ--and takes more processing power.  I saw a roughly 25% performance increase in a few (somewhat unscientific tests) of using three different stops with all couplers versus three duplicated stops with all couplers, but it seemed consistent enough to merit investigation.

I would guess that adding some sort of if-then to the code that checks to see if a pipe is already in use rather than allowing the same note to run through individual divisional tuning and then layering could immediately and simply save on performance, perhaps allowing for Full Organ--at least more easily or on more occasions--that the present software allows.

Spotify: https://open.spotify.com/artist/2xHiPcCsm29R12HX4eXd4J
Pianoteq Studio & Organteq
Casio GP300 & Custom organ console

Re: Hauptwerk and Organteq comparison?

Tutti wrote:

Same observation here. Tutti with all couplers engaged is impossible, but then my music PC is quite old and underpowered. For now, I can play tutti not coupling anything just fine. Probably future versions could make use of CPU power more efficiently? Down the road it should be an i7 or i9 I suppose.

Apart from the problems reported by tmyoung which can only solve Modartt, the computing power must be measured in GFLOP but it is the specific physical model that loads the CPU and can be optimized with parallel processing and with optimized instructions found in the new processors (AVX, AVX2, AVX-512, FMAx). The upcoming Hauptwerk version 5 will also use these instructions!
Today's processors have a power of hundreds of GFLOP but with current developments we are beyond the TFLOP (in particular the AMD Zen 2 series up to 64 cores) without considering accelerator cards like Xeon Phi.

For an equal physical modelling comparison, Viscount UNICO series uses up to 8 SHARC processors for a maximum power of about 12 GFLOP. A SHARC DSP are used in embedded systems but are a hundred times less powerful than a current Intel / AMD processor.

With the old Intel CPU you could have these performances:
Processor                             Brief Spec                                Linpack (GFLOPS)
Dual Xeon E5 2687W              16 cores @ 3.2GHz AVX              345
Core i7 5930K (Haswell E)      6 cores @ 3.5GHz AVX2             289
Dual Xeon E5 2650                 16 cores @ 2.0GHz AVX              262
Core i7 4770K (Haswell)          4 cores @ 3.5GHz AVX2            182

Re: Hauptwerk and Organteq comparison?

tmyoung wrote:
Tutti wrote:

Same observation here. Tutti with all couplers engaged is impossible, but then my music PC is quite old and underpowered. For now, I can play tutti not coupling anything just fine. Probably future versions could make use of CPU power more efficiently? Down the road it should be an i7 or i9 I suppose.

I've noticed that duplicate instances of stops stack, like the Trompette 8' which is part of the Pedal, Grand Organ, and Recit divisionals.  On a regular organ, the ranks for this stop would only be in one divisional--likely the Grand Organ, while having optional redundant stop knobs on the other manuals, which means that when a note is played on another of the manuals or via a coupler, the organ still will only make the sound once.

In this version of Organteq, the sound is simulated multiple times through different tuning settings and layered, which both makes the sound a little louder--which wouldn't happen with a real organ--and takes more processing power.  I saw a roughly 25% performance increase in a few (somewhat unscientific tests) of using three different stops with all couplers versus three duplicated stops with all couplers, but it seemed consistent enough to merit investigation.

I would guess that adding some sort of if-then to the code that checks to see if a pipe is already in use rather than allowing the same note to run through individual divisional tuning and then layering could immediately and simply save on performance, perhaps allowing for Full Organ--at least more easily or on more occasions--that the present software allows.

Great observation. It's too early for me to think logically, but if I just selected all the unique stops per each manual and then couple, it should rectify the problem of layering!

Re: Hauptwerk and Organteq comparison?

aile wrote:
Tutti wrote:

Same observation here. Tutti with all couplers engaged is impossible, but then my music PC is quite old and underpowered. For now, I can play tutti not coupling anything just fine. Probably future versions could make use of CPU power more efficiently? Down the road it should be an i7 or i9 I suppose.

Apart from the problems reported by tmyoung which can only solve Modartt, the computing power must be measured in GFLOP but it is the specific physical model that loads the CPU and can be optimized with parallel processing and with optimized instructions found in the new processors (AVX, AVX2, AVX-512, FMAx). The upcoming Hauptwerk version 5 will also use these instructions!
Today's processors have a power of hundreds of GFLOP but with current developments we are beyond the TFLOP (in particular the AMD Zen 2 series up to 64 cores) without considering accelerator cards like Xeon Phi.

For an equal physical modelling comparison, Viscount UNICO series uses up to 8 SHARC processors for a maximum power of about 12 GFLOP. A SHARC DSP are used in embedded systems but are a hundred times less powerful than a current Intel / AMD processor.

With the old Intel CPU you could have these performances:
Processor                             Brief Spec                                Linpack (GFLOPS)
Dual Xeon E5 2687W              16 cores @ 3.2GHz AVX              345
Core i7 5930K (Haswell E)      6 cores @ 3.5GHz AVX2             289
Dual Xeon E5 2650                 16 cores @ 2.0GHz AVX              262
Core i7 4770K (Haswell)          4 cores @ 3.5GHz AVX2            182


Interesting. In other words, if tmyoung is onto something there that can and probably will be fixed in a future update, my modest AMD A8 Fusion (4 cores at 2.9 GHz) could last another round or two of updates (HW 5, OrganTeq).

I do not need large dispositions to toy with. So far, the free Paramount 3/10 has lasted my TO musings, and the OrganTeq model is perfect for my modest practice duties. Love the sound and quick boot. I toyed with GO and the excellent free sample sets, but never liked the long boot times of any of the sample sets. Also, OrganTeq might be finally the solution to use it at church (old digital Baldwin) using my MacBook (i5).

Re: Hauptwerk and Organteq comparison?

Tutti wrote:

Great observation. It's too early for me to think logically, but if I just selected all the unique stops per each manual and then couple, it should rectify the problem of layering!

Well, it's my turn to claim it's too early for me to think, because I should have just thought of trying your idea of full organ with couplers and turning off duplicate stops rather than my single-stop tests...

With the duplicate stops and polyphony set to 96, my dual Xenon E5405 processors (a decent but 10-year old system) couldn't do full organ with couplers on a six-voice fugue, but with duplicate stops removed, I could go to up to 160 polyphony on full organ and couplers with no audio loss/performance issues (there was some beating on long pedal tones but that had more to do with the polyphony limit--I suspect--but none of the "crunching" artefacts from processor overload).

Based on this test--especially if you can manage a little extra ASIO delay if the processor is really old--you should be able to run Organteq on anything even with a moderate polyphony limit.

Spotify: https://open.spotify.com/artist/2xHiPcCsm29R12HX4eXd4J
Pianoteq Studio & Organteq
Casio GP300 & Custom organ console

Re: Hauptwerk and Organteq comparison?

tmyoung wrote:

Based on this test--especially if you can manage a little extra ASIO delay if the processor is really old--you should be able to run Organteq on anything even with a moderate polyphony limit.

A quality audio board and good ASIO driver apparently will also help the CPU load. I just played around with the two boards (both consumer grade), and the Xonar definitely is helping to the point that OrganTeq is almost glitch free now. Not completely, but to the point I can neglect a CPU upgrade for some time to come.

Re: Hauptwerk and Organteq comparison?

aile wrote:
Tutti wrote:

Same observation here. Tutti with all couplers engaged is impossible, but then my music PC is quite old and underpowered. For now, I can play tutti not coupling anything just fine. Probably future versions could make use of CPU power more efficiently? Down the road it should be an i7 or i9 I suppose.

Apart from the problems reported by tmyoung which can only solve Modartt, the computing power must be measured in GFLOP but it is the specific physical model that loads the CPU and can be optimized with parallel processing and with optimized instructions found in the new processors (AVX, AVX2, AVX-512, FMAx). The upcoming Hauptwerk version 5 will also use these instructions!
Today's processors have a power of hundreds of GFLOP but with current developments we are beyond the TFLOP (in particular the AMD Zen 2 series up to 64 cores) without considering accelerator cards like Xeon Phi.

For an equal physical modelling comparison, Viscount UNICO series uses up to 8 SHARC processors for a maximum power of about 12 GFLOP. A SHARC DSP are used in embedded systems but are a hundred times less powerful than a current Intel / AMD processor.

With the old Intel CPU you could have these performances:
Processor                             Brief Spec                                Linpack (GFLOPS)
Dual Xeon E5 2687W              16 cores @ 3.2GHz AVX              345
Core i7 5930K (Haswell E)      6 cores @ 3.5GHz AVX2             289
Dual Xeon E5 2650                 16 cores @ 2.0GHz AVX              262
Core i7 4770K (Haswell)          4 cores @ 3.5GHz AVX2            182

With my laptop with a 9900k i9 core processor, and without limiting the polyphony, I typically use between 20 and 22% of CPU resources. at the extreme an effective polyphony displayed around 350 I stayed below 40% of the resources. (but 70% with all games enabled)

Bruno

Re: Hauptwerk and Organteq comparison?

bm wrote:
aile wrote:
Tutti wrote:

Same observation here. Tutti with all couplers engaged is impossible, but then my music PC is quite old and underpowered. For now, I can play tutti not coupling anything just fine. Probably future versions could make use of CPU power more efficiently? Down the road it should be an i7 or i9 I suppose.

Apart from the problems reported by tmyoung which can only solve Modartt, the computing power must be measured in GFLOP but it is the specific physical model that loads the CPU and can be optimized with parallel processing and with optimized instructions found in the new processors (AVX, AVX2, AVX-512, FMAx). The upcoming Hauptwerk version 5 will also use these instructions!
Today's processors have a power of hundreds of GFLOP but with current developments we are beyond the TFLOP (in particular the AMD Zen 2 series up to 64 cores) without considering accelerator cards like Xeon Phi.

For an equal physical modelling comparison, Viscount UNICO series uses up to 8 SHARC processors for a maximum power of about 12 GFLOP. A SHARC DSP are used in embedded systems but are a hundred times less powerful than a current Intel / AMD processor.

With the old Intel CPU you could have these performances:
Processor                             Brief Spec                                Linpack (GFLOPS)
Dual Xeon E5 2687W              16 cores @ 3.2GHz AVX              345
Core i7 5930K (Haswell E)      6 cores @ 3.5GHz AVX2             289
Dual Xeon E5 2650                 16 cores @ 2.0GHz AVX              262
Core i7 4770K (Haswell)          4 cores @ 3.5GHz AVX2            182

With my laptop with a 9900k i9 core processor, and without limiting the polyphony, I typically use between 20 and 22% of CPU resources. at the extreme an effective polyphony displayed around 350 I stayed below 40% of the resources. (but 70% with all games enabled)

Bruno


...thanks for the information, Modartt has obviously sized the computing power with modern CPUs. Keep in mind that Intel i9 9900k is a powerful 8 cores 5GHz cpu with AVX2 support.

It may be more practical to have a high-end CPU comparison with the following link:
https://www.cpubenchmark.net/high_end_cpus.html

Re: Hauptwerk and Organteq comparison?

I don't think Organteq is using AVX2 (or even AVX) instructions - but Roman might want to reply to this one. But having very fast cores (rather than many more slower cores) is always going to be better for all virtual instruments. In other words, if you have 8 cores closer to 5 GHz, it is in most cases going to be better than 16 cores at closer to 3 GHz.


If you want to use full polyphony of 512 voices, indeed you want to have a beefy CPU. My i7-6700 is running at 4.5 GHz, and with multicore enabled it starts choking at around 300-320 voices... So yeah 9900K sounds like the ticket here.

Last edited by EvilDragon (03-12-2019 16:48)
Hard work and guts!

Re: Hauptwerk and Organteq comparison?

EvilDragon wrote:

If you want to use full polyphony of 512 voices, indeed you want to have a beefy CPU. My i7-6700 is running at 4.5 GHz, and with multicore enabled it starts choking at around 300-320 voices... So yeah 9900K sounds like the ticket here.

My system does fine with 512 polyphony.  Intel Core i7-6900K @ 3.2 GHz, 8 cores.  Good ASIO Sound-Card by RME.

Re: Hauptwerk and Organteq comparison?

RME UFX+ over here too. 48 kHz at 128 samples buffer size. Maybe it's because of the buffer size. If I go to 256 it should be fine. Which buffer size are you on?

EDIT: At 256 samples buffer it chokes at ~400 voices. But I appreciate lower latency, so...

Last edited by EvilDragon (03-12-2019 18:50)
Hard work and guts!

Re: Hauptwerk and Organteq comparison?

EvilDragon wrote:

RME UFX+ over here too. 48 kHz at 128 samples buffer size. Maybe it's because of the buffer size. If I go to 256 it should be fine. Which buffer size are you on?

EDIT: At 256 samples buffer it chokes at ~400 voices. But I appreciate lower latency, so...

I tried that and my Duo-Capture EX instead of the built-in. Same result for me...
I suppose you know that pipe organs (especially mechanical ones) have huge latency so a larger buffer is actually more realistic here.

Re: Hauptwerk and Organteq comparison?

Yes I know they have larger latency, which is a thing I don't like about real organs

Hard work and guts!

Re: Hauptwerk and Organteq comparison?

EvilDragon wrote:

If I go to 256 it should be fine. Which buffer size are you on?

EDIT: At 256 samples buffer it chokes at ~400 voices. But I appreciate lower latency, so...

Running at Buffer 256, nominal latency 5.3 ms which feels about right for Pipe Organ.  The extra Cores (8 in my system) probably help a lot.

Re: Hauptwerk and Organteq comparison?

OrganTeq 1.0.0 puts 38 seconds to start on AMD-A6-3610 1.80Ghz and 6Gb RAM

Re: Hauptwerk and Organteq comparison?

OrganoPleno wrote:
EvilDragon wrote:

If I go to 256 it should be fine. Which buffer size are you on?

EDIT: At 256 samples buffer it chokes at ~400 voices. But I appreciate lower latency, so...

Running at Buffer 256, nominal latency 5.3 ms which feels about right for Pipe Organ.  The extra Cores (8 in my system) probably help a lot.

For sure, you have twice as many cores that I do! That would definitely help with Organteq.



OST999 wrote:

OrganTeq 1.0.0 puts 38 seconds to start on AMD-A6-3610 1.80Ghz and 6Gb RAM


Yep that's a slow CPU.

Last edited by EvilDragon (03-12-2019 23:40)
Hard work and guts!

Re: Hauptwerk and Organteq comparison?

EvilDragon wrote:
OST999 wrote:

OrganTeq 1.0.0 puts 38 seconds to start on AMD-A6-3610 1.80Ghz and 6Gb RAM


Yep that's a slow CPU.

I forgot to specify, he has 4 cores, but it's same with my AMD-A10 at 3.50 Ghz and 4 cores, 32Gb RAM, and the 2 with Windows 7 SP1
Nevertheless, it started much faster the first time....

Last edited by OST999 (04-12-2019 00:00)

Re: Hauptwerk and Organteq comparison?

bm wrote:

With my laptop with a 9900k i9 core processor, and without limiting the polyphony, I typically use between 20 and 22% of CPU resources. at the extreme an effective polyphony displayed around 350 I stayed below 40% of the resources. (but 70% with all games enabled)
Bruno

In fact, with all stops enabled, the maximum polyphony of 512 is quickly reached without audio problem (playing quickly + putting the 2 elbows on the keyboard - not very comfortable nor harmonious - ...) Organteq indicates a consumption of resources between 60 to 67%, the task manager 32%, with a CPU frequency (average?) Of 4,67Ghz [my laptop is configured in performance mode, but in fact constrained by its power supply and the automatic mode of management of the fan, that I do not wish to see too noisy, voluntarily not forced to the maximum, and deliberately no configuration of CPU overclocking mode, the maximum theoretical frequency of 5Ghz processor 2 cores is not reached here] On 16 cores (8 cores / 16 simultaneous threads) 14 have a significant activity, of which 1 to 80%, 6 to 50% approximately) By playing "normally" (even with all stops enabled) the fans remain silent, 8% of the CPU power is requested according to task manager, rarely more than 17% of resources indicated by Organteq)
Bruno

Re: Hauptwerk and Organteq comparison?

It looks like everyone agrees that OrganTeq requires the highest performance CPU you can afford!

Here's a good place you can go to compare the CPU you have vs. what's available, and at what price:

https://www.cpubenchmark.net/high_end_cpus.html

For example, the Intel Core i9-7940X @ 3.10GHz (rated at 25,426) is a powerful Intel solution, costing US$909.99.

Or, for best value, choose an AMD Ryzen 9 3900X is rated at 31,916 and costs US$549.99

These are 4-5 times more powerful than your average i5 or i7 CPU.

For best performance, use a top rated CPU in a Linux system!

Re: Hauptwerk and Organteq comparison?

For a long time, I've always wanted the AMD Ryzen 9 3900X but it's expensive

Re: Hauptwerk and Organteq comparison?

Not as expensive as Intel, at least.

Hard work and guts!

Re: Hauptwerk and Organteq comparison?

EvilDragon wrote:

Not as expensive as Intel, at least.

If I had to make a better compromise on the price of the processor, I would certainly be interested in the Rizen 5 3600:
see https://www.cpubenchmark.net/cpu.php?cp...mp;id=3481
* Performances already very high ON 1 ONLY thread (essential for Organteq / Pianoteq because there is always a logical core which works much more than the others and is in fact the limiting factor)
* Number of logical core sufficient (here 12) including 6 physical, I found that Organteq (idem Pianoteq) does not use all the cores with 16 logical core for example.
* TDP not too important (here 65W): can find a motherboard at a more moderate price, and especially to run the fans slower (.. it is still interesting to hear a little more the instrument rather that the machine, especially with open headphones ...)

// In addition, I just used my 2nd Organteq slot on a smaller Dell laptop recently purchased, which has an Intel Core i5 9300H processor (8 logical core, 4 physical)
link: https://www.cpubenchmark.net/cpu.php?cp...mp;id=3448 .The default max polyphony: 160, works perfectly (use on the same external dac and my keyboard casio GP500) mode "tutti" (all stops) I can put elbows on the keyboard without problems without saturating the processor. The maximum level of polyphony that can be used (except bends on the keyboard) is 256, a polyphony that can actually be achieved with about 90% of CPU resources used with Organteq. In practice the consumption of resources is between 25 and 40% according to the indicator of Organteq playing "normally", in "tutti" mode. (My Dell laptop which originally had 8GB of RAM had been upgraded to 16GB for a few dozen euros of RAM bar - probably useless for Organteq)

Bruno

Last edited by bm (07-12-2019 07:49)

Re: Hauptwerk and Organteq comparison?

I have a question for Organteq Mac users with cpus having more than 4 cores: How many extra cores (or hyperthreads) are significantly used when pushing Organteq to high polyphony?

I'm asking because I installed it also on my recent MacBook Air 2019 (that I bought for simpler use, not for power) and I was surprised that on that hyperthreaded dual core i5, all 4 logical threads were fully and equally used, and Organteq works very well on it, while on my older MacPro 2013 only the 4 main cores are used (no hyperthreading).

Maybe there are some new instructions that were added in newer cpus or a difference in OS. My MacPro is still on OSX 10.9.5 while the new MacBook Air is under Mojave. Anyway, I just realized how much progress was done in the last 5-6 years to have a lowly laptop i5 be about only half as powerful as a 2013 Xeon...

This is also to say that while it is possible to push cpu usage to very high level with Organteq, by no mean is it limited to very powerful system for normal use.

Last edited by Gilles (07-12-2019 16:42)

Re: Hauptwerk and Organteq comparison?

EvilDragon wrote:
OST999 wrote:

OrganTeq 1.0.0 puts 38 seconds to start on AMD-A6-3610 1.80Ghz and 6Gb RAM


Yep that's a slow CPU.

No, apparently, it's a bug and they will probably be able to fix it !
Again same bug in 1.0.1...