or may be its time for some few Re dev. on Reason to take care seriously DSP/CPU optimisation ..OoMarco Raaphorst wrote:Not sure what is going on but Reason is performing worse than any other DAW. Much worse. Hope the Props take it very serious this time. I am having problems with guitar and low latency for a long time in Reason. Different problem probably.joeyluck wrote:SynapseAudio wrote:We use the exact same codebase for VST/RE so it doesn't look like we can do anything on our end. The Props are looking into it, however.doctecazoid wrote: hope in a future version they are able to optimize the code so the CPU takes less of a hit.I've noticed that my DSP meter jumps up after notes are released; after silence. Sometimes jumping to three bars from one or two.Marco Raaphorst wrote:Yes Reason has issues compared to the AU/VST unit. I compared both version and it was shocking: http://melodiefabriek.com/blog/synapse- ... versus-au/damasio wrote:I am currently demoying this synth and how much i am impressed with its sound,
the CPU load is unbearable on my mac 8 core. You guys really need work this out because cpu load is waaay out of control.
One instance + redrum and some reverb on 44.1 rips my cpu to shreds.
The Props also responded because they admitted this is strange behaviour. I am not sure what they will do but this RE related performance issue is a serious shop stopper at the moment.
Maybe that has something to do with whatever is going on?
Synapse The Legend Synthesizer is in the Shop
- frog974new
- Posts: 352
- Joined: 16 Jan 2015
- Contact:
Hi Marco,Marco Raaphorst wrote:Yes Reason has issues compared to the AU/VST unit. I compared both version and it was shocking: http://melodiefabriek.com/blog/synapse- ... versus-au/damasio wrote:I am currently demoying this synth and how much i am impressed with its sound,
the CPU load is unbearable on my mac 8 core. You guys really need work this out because cpu load is waaay out of control.
One instance + redrum and some reverb on 44.1 rips my cpu to shreds.
The Props also responded because they admitted this is strange behaviour. I am not sure what they will do but this RE related performance issue is a serious shop stopper at the moment.
is it possible, that one reason for the difference in cpu usage between Reason and other DAW's is that Reason only uses 3 out of 4 CPU cores?
In case that I'm right and the other DAW's use all 4 cores, then it should be expected that there is a big difference.
In my opinion Propellerhead should work on a solution, that Reason can use all 4 cores, that would be a major progress.
There is one core reserved for the GUI. That is a good idea in general, but maybe Hyperthreading can be used there rather than one complete core.riemac wrote:Hi Marco,Marco Raaphorst wrote:Yes Reason has issues compared to the AU/VST unit. I compared both version and it was shocking: http://melodiefabriek.com/blog/synapse- ... versus-au/damasio wrote:I am currently demoying this synth and how much i am impressed with its sound,
the CPU load is unbearable on my mac 8 core. You guys really need work this out because cpu load is waaay out of control.
One instance + redrum and some reverb on 44.1 rips my cpu to shreds.
The Props also responded because they admitted this is strange behaviour. I am not sure what they will do but this RE related performance issue is a serious shop stopper at the moment.
is it possible, that one reason for the difference in cpu usage between Reason and other DAW's is that Reason only uses 3 out of 4 CPU cores?
In case that I'm right and the other DAW's use all 4 cores, then it should be expected that there is a big difference.
In my opinion Propellerhead should work on a solution, that Reason can use all 4 cores, that would be a major progress.
Reason12, Win10
I can understand that it is a good idea to reserve CPU power for the GUI, but my 4. core is most of the time only used 5% - 10%.Loque wrote: There is one core reserved for the GUI. That is a good idea in general, but maybe Hyperthreading can be used there rather than one complete core.
There is a lot of CPU power wasted and I would be glad if I could allocate this in the preferences of Reason for audio calculation.
Reason allocates all cores as required. If you receive the CPU warning, it may be a result of a CPU peak due to bad coding or bugs, often raised while switching patches. Just turn the warning off, if you have more than one core.riemac wrote:I can understand that it is a good idea to reserve CPU power for the GUI, but my 4. core is most of the time only used 5% - 10%.Loque wrote: There is one core reserved for the GUI. That is a good idea in general, but maybe Hyperthreading can be used there rather than one complete core.
There is a lot of CPU power wasted and I would be glad if I could allocate this in the preferences of Reason for audio calculation.
Reason12, Win10
When I switch off the CPU limit in the preferences and run a too cpu heavy project, the playback stops and I can see in the cpu diagnostic of the taskmanager, that the 4th core still isn't used at all. All three other physical cores are at their maximum.Loque wrote:Reason allocates all cores as required. If you receive the CPU warning, it may be a result of a CPU peak due to bad coding or bugs, often raised while switching patches. Just turn the warning off, if you have more than one core.riemac wrote:I can understand that it is a good idea to reserve CPU power for the GUI, but my 4. core is most of the time only used 5% - 10%.Loque wrote: There is one core reserved for the GUI. That is a good idea in general, but maybe Hyperthreading can be used there rather than one complete core.
There is a lot of CPU power wasted and I would be glad if I could allocate this in the preferences of Reason for audio calculation.
In my opinion this is the proof, that Reason doesn't allocate all physical cores, but uses only 3 of the 4 physical cores.
- Marco Raaphorst
- Posts: 2504
- Joined: 22 Jan 2015
- Location: The Hague, The Netherlands
- Contact:
I had a quick view on my MacBook Pro. I see all 8 cores being used just like in Live.riemac wrote:Hi Marco,Marco Raaphorst wrote:Yes Reason has issues compared to the AU/VST unit. I compared both version and it was shocking: http://melodiefabriek.com/blog/synapse- ... versus-au/damasio wrote:I am currently demoying this synth and how much i am impressed with its sound,
the CPU load is unbearable on my mac 8 core. You guys really need work this out because cpu load is waaay out of control.
One instance + redrum and some reverb on 44.1 rips my cpu to shreds.
The Props also responded because they admitted this is strange behaviour. I am not sure what they will do but this RE related performance issue is a serious shop stopper at the moment.
is it possible, that one reason for the difference in cpu usage between Reason and other DAW's is that Reason only uses 3 out of 4 CPU cores?
In case that I'm right and the other DAW's use all 4 cores, then it should be expected that there is a big difference.
In my opinion Propellerhead should work on a solution, that Reason can use all 4 cores, that would be a major progress.
Thanks. That is strange, but maybe Reason behaves different on a MAC.Marco Raaphorst wrote:I had a quick view on my MacBook Pro. I see all 8 cores being used just like in Live.riemac wrote:Hi Marco,Marco Raaphorst wrote:Yes Reason has issues compared to the AU/VST unit. I compared both version and it was shocking: http://melodiefabriek.com/blog/synapse- ... versus-au/damasio wrote:I am currently demoying this synth and how much i am impressed with its sound,
the CPU load is unbearable on my mac 8 core. You guys really need work this out because cpu load is waaay out of control.
One instance + redrum and some reverb on 44.1 rips my cpu to shreds.
The Props also responded because they admitted this is strange behaviour. I am not sure what they will do but this RE related performance issue is a serious shop stopper at the moment.
is it possible, that one reason for the difference in cpu usage between Reason and other DAW's is that Reason only uses 3 out of 4 CPU cores?
In case that I'm right and the other DAW's use all 4 cores, then it should be expected that there is a big difference.
In my opinion Propellerhead should work on a solution, that Reason can use all 4 cores, that would be a major progress.
I've been using this for the last 45 min. Don't hear anything that I can't do with Thor alone or Thor with Audiomatic or a couple other stock devices. Running and i5 with 1 DSP bar in mono mode and 2 in poly mode. Thor doesn't even register on the DSP meter. Already re-created some of Legend's patches on thor...even the ones with amp gain/overdrive. I am a Moog owner/user as well since 1986 (Source and Polymoog,) so I'm pretty good at the old analog game. Was not impressed by Viking when I trialed that either.
Can you pls provide some examples or compareable patches? Preferable as patch/combinator files. If you provide samples, please provide also very low and very high tones.bpmorton wrote:I've been using this for the last 45 min. Don't hear anything that I can't do with Thor alone or Thor with Audiomatic or a couple other stock devices. Running and i5 with 1 DSP bar in mono mode and 2 in poly mode. Thor doesn't even register on the DSP meter. Already re-created some of Legend's patches on thor...even the ones with amp gain/overdrive. I am a Moog owner/user as well since 1986 (Source and Polymoog,) so I'm pretty good at the old analog game. Was not impressed by Viking when I trialed that either.
Reason12, Win10
- SynapseAudio
- RE Developer
- Posts: 59
- Joined: 22 Jan 2015
Would be interested in that too, recreate the default patch of The Legend exactly please, and some of the vintage ones too
And btw the Polymoog is a string synth with divide-down circuitry. It's not even close in concept or sound.
Richard
And btw the Polymoog is a string synth with divide-down circuitry. It's not even close in concept or sound.
Richard
let me be clear here: I'm not picking a fight and this should not devolve into the "kong look no further" thread. I'm not trying to dissuade folks from buying this RE. It has a great interface and is 34x cheaper than buying a mini, vintage or new. There will be some subtleties that can't be done with Thor but may not be noticeable in a mix . So far here are two patches. The default Bass Carpet patch and the Cinematic Strings patch.SynapseAudio wrote:Would be interested in that too, recreate the default patch of The Legend exactly please, and some of the vintage ones too
And btw the Polymoog is a string synth with divide-down circuitry. It's not even close in concept or sound.
Richard
- Attachments
-
- bpmorton777 Legend clone.zip
- (4.41 KiB) Downloaded 132 times
Im loving this synth. It has that extra detail that you not get from most other Moog emulations. Thor can do a lot of sounds of course but it doesn't do everything just as well as other synths might. So to me its kind of silly to say that Thor can easily do the sound of The Legend. Which it cant. If you look at it like "you wont hear it in the mix" then you're not that passionate about sound and thats oke but many others are. You could post that answer in every thread about a new synth...
On the Cinematic Strings, you are missing the oscillator (and filter) drift, the 2nd filter without keytracking and the pitch was slightly down a few cents, a bit more wet on the reverb and a few other tweaks.. Sounds great though not the same rich sound, but after a few mods i made its very close!bpmorton wrote:
let me be clear here: I'm not picking a fight and this should not devolve into the "kong look no further" thread. I'm not trying to dissuade folks from buying this RE. It has a great interface and is 34x cheaper than buying a mini, vintage or new. There will be some subtleties that can't be done with Thor but may not be noticeable in a mix . So far here are two patches. The default Bass Carpet patch and the Cinematic Strings patch.
Here is the updated patch i made:
https://www.dropbox.com/s/sddj9yrztp89b ... 9.cmb?dl=1
with the extra LFO's doing drift the sound is has a little too much chorus. That is why I left those out. Trust your ears. Other legend patches that I've demoed have the tuning dead center on all VCO's and let the chorusing sound from the drift. with it that subtle you gotta pick one or the other: tuning drift or detuned VCO's. Much easier to just grab the detune knob.dana wrote:On the Cinematic Strings, you are missing the oscillator (and filter) drift, the 2nd filter without keytracking and the pitch was slightly down a few cents, a bit more wet on the reverb and a few other tweaks.. Sounds great though not the same rich sound, but after a few mods i made its very close!bpmorton wrote:
let me be clear here: I'm not picking a fight and this should not devolve into the "kong look no further" thread. I'm not trying to dissuade folks from buying this RE. It has a great interface and is 34x cheaper than buying a mini, vintage or new. There will be some subtleties that can't be done with Thor but may not be noticeable in a mix . So far here are two patches. The default Bass Carpet patch and the Cinematic Strings patch.
Here is the updated patch i made:
https://www.dropbox.com/s/sddj9yrztp89b ... 9.cmb?dl=1
here is another Patch: French KS. This one is more difficult because of the mix of noise/VCO3 modulating the pitch of all VCO's. I had to use a second Thor, a mixer and a spider to do this.
- Attachments
-
- French KS Clone.zip
- (2.45 KiB) Downloaded 103 times
- Marco Raaphorst
- Posts: 2504
- Joined: 22 Jan 2015
- Location: The Hague, The Netherlands
- Contact:
Great patches!bpmorton wrote:let me be clear here: I'm not picking a fight and this should not devolve into the "kong look no further" thread. I'm not trying to dissuade folks from buying this RE. It has a great interface and is 34x cheaper than buying a mini, vintage or new. There will be some subtleties that can't be done with Thor but may not be noticeable in a mix . So far here are two patches. The default Bass Carpet patch and the Cinematic Strings patch.SynapseAudio wrote:Would be interested in that too, recreate the default patch of The Legend exactly please, and some of the vintage ones too
And btw the Polymoog is a string synth with divide-down circuitry. It's not even close in concept or sound.
Richard
To be fair, that thread was fucking epic. A spectacular rainbow of human idiocybpmorton wrote:this should not devolve into the "kong look no further" thread.
- SynapseAudio
- RE Developer
- Posts: 59
- Joined: 22 Jan 2015
The Cinematic Strings impersonation is too far away from the original for my taste, it's more like a new patch, one that is great though!
The Bass Carpet is close and nicely programmed, but when actually playing it, it sounds to me like a $100 guitar from China trying to impersonate a Gibson Les Paul. It is similar, just like two guitars would both sound like a "guitar", but something is severely lacking in sound quality in the clone. That's of course my subjective view but in any case it's quite far from being identical.
The ultimate test is if you do sequences as in our comparison vid:
That is, if you make an entire sequence that cannot be distinguished from either of those clips, Original or Legend. This is yet harder than getting close to the sound alone, because pb/glide, tuning and time-varying effects are added on top of the already substantial differences.
Richard
The Bass Carpet is close and nicely programmed, but when actually playing it, it sounds to me like a $100 guitar from China trying to impersonate a Gibson Les Paul. It is similar, just like two guitars would both sound like a "guitar", but something is severely lacking in sound quality in the clone. That's of course my subjective view but in any case it's quite far from being identical.
The ultimate test is if you do sequences as in our comparison vid:
That is, if you make an entire sequence that cannot be distinguished from either of those clips, Original or Legend. This is yet harder than getting close to the sound alone, because pb/glide, tuning and time-varying effects are added on top of the already substantial differences.
Richard
These sound good and are great patch emulationsbpmorton wrote:let me be clear here: I'm not picking a fight and this should not devolve into the "kong look no further" thread. I'm not trying to dissuade folks from buying this RE. It has a great interface and is 34x cheaper than buying a mini, vintage or new. There will be some subtleties that can't be done with Thor but may not be noticeable in a mix . So far here are two patches. The default Bass Carpet patch and the Cinematic Strings patch.SynapseAudio wrote:Would be interested in that too, recreate the default patch of The Legend exactly please, and some of the vintage ones too
And btw the Polymoog is a string synth with divide-down circuitry. It's not even close in concept or sound.
Richard
But the differences are apparent to my ears and much of the character lacking.
And I'm sure yourself or somebody can spend more time emulating these same patches to get them a bit closer while using a Combinator, etc.
But I think the main take away is that these lack the interface (as you mention), the character, and the behavior of parameter adjustment.
For instance, filter adjustments are very different. Let alone all the other parameters and how they interact. So in other words, these serve as good, static patch emulations; meaning parameters frozen. Similarly I'm sure we could see someone else chime in soon talking about a Moog sample library and then someone else replying, "That's great until you want to edit the sound."
Big thumbs up in terms of sound design. You've got talent!
But much difference between saying, "Here's some Legend-sounding patches" vs. "Here's a Legend"
Or is it "Here's a The Legend?"
That was interesting to see/hear - I don't consider myself to have amazing ears, but could definitely discern subtle differences between most of the examples there. Which doesn't mean the hardware necessarily sounded better than the software (except the last example, where I thought the hardware had just a touch more heft & presence to it)SynapseAudio wrote:
The ultimate test is if you do sequences as in our comparison vid:
- SynapseAudio
- RE Developer
- Posts: 59
- Joined: 22 Jan 2015
Try those blind tests if you like:dvdrtldg wrote:That was interesting to see/hear - I don't consider myself to have amazing ears, but could definitely discern subtle differences between most of the examples there. Which doesn't mean the hardware necessarily sounded better than the software (except the last example, where I thought the hardware had just a touch more heft & presence to it)
http://www.synapse-audio.com/legend/LegendBlindTest.zip
The clips are similar to those in the video, but the order is random. So far we have no evidence that anyone can spot the hardware reliably by ear, and it has been attempted by quite a few producers so far. I posted the clips over at GS so the solution is known, but if you haven't seen them yet, feel free to give it a try
Subtle differences are there for sure, if for no other reason than that oscillators are free-running- so every time you press a key, the transients look and sound different. This is also why the graphical representation of the audio looks different in the comparison video if you look closely.
Richard
Very impressive results and some of the most convincing analog hardware mimicry I've heard up to this point. It consistently maintains its character remarkably well, even when pushed to more extreme settings and I like the little bit of unpredictability with every key press.SynapseAudio wrote:Try those blind tests if you like:dvdrtldg wrote:That was interesting to see/hear - I don't consider myself to have amazing ears, but could definitely discern subtle differences between most of the examples there. Which doesn't mean the hardware necessarily sounded better than the software (except the last example, where I thought the hardware had just a touch more heft & presence to it)
http://www.synapse-audio.com/legend/LegendBlindTest.zip
The clips are similar to those in the video, but the order is random. So far we have no evidence that anyone can spot the hardware reliably by ear, and it has been attempted by quite a few producers so far. I posted the clips over at GS so the solution is known, but if you haven't seen them yet, feel free to give it a try
Subtle differences are there for sure, if for no other reason than that oscillators are free-running- so every time you press a key, the transients look and sound different. This is also why the graphical representation of the audio looks different in the comparison video if you look closely.
Richard
Music is nothing else but wild sounds civilized into time and tune.
The cover patches are not bad, well done. But some of the patches of The Legend are absolutly unique and the response of my FX to this synth is pretty amazing. The filter and the reverb are actually unique and gonna fill my arsenal as well and just they are worth 50€. If all filters and reverbs the same, i did not had bought nearly all of them
I am pretty sure, lots of quite simple patches can be reproduced by other synths, at least closely. We had in some thread a dicussion about simple saws and the different frequencey-spectrums the different synths created. I will have a deeper look into the "Lo" oscilator and what he does, because this seems to be not so easy to be reproduced by LFOs or synced oscilators...
I am pretty sure, lots of quite simple patches can be reproduced by other synths, at least closely. We had in some thread a dicussion about simple saws and the different frequencey-spectrums the different synths created. I will have a deeper look into the "Lo" oscilator and what he does, because this seems to be not so easy to be reproduced by LFOs or synced oscilators...
Reason12, Win10
The idea behind such REs is that you don't need waste time in recreating sounds or searching for good patches. And the results and the handling of the Legend are much closer.Don't hear anything that I can't do with Thor alone or Thor with Audiomatic or a couple other stock devices.
-
- Information
-
Who is online
Users browsing this forum: No registered users and 9 guests