SynthMaster 2.0.3.3 - Post bugs here!

[Ingonator]

Looks like 1.) is not fixed, still two overlaying knobs.

In beta 21 I could still hear crackles even if they are less, mostly in the patches from rsmus7. Could anyone here confirm this or is it only a problem of my system? I tried the patches with the standalone version. Most patches (all patches without unison) don't cause problems so i guess it's not just a problem with my system.
Maybe rsmus7 could check this himself if he finds some time.

Apparently I forgot to copy over the interface.xml file to the latest build. Anyhow will do that this weekend.

Regarding crackles, they are the result of "hard clipping" I believe, so not a bug technically, just due to excessive volume. If preset volumes were lowered, it should go away, and for unison I already took care of that by scaling down the volume when unison voices are turned on.

Until the end of next week my concentration will still be on FX: Particularly Reverb & Phaser. The phaser is very easy to implement, will have between 1-8 notches... Might add feedback parameter as well.

For the reverb, I'll make it more visually editable and add lo/hishelf EQs to the late reverberation part. Also will rewrite the early reflection part so it sounds more "traditional".

Please check the Update of my post. I found some new strange things.

UPDATE: I found that the high CPU use in rsmus7's pad sounds is caused by high unison settings. The "PAD coldsmoother III SM" patch uses 8 voices unison in Layer 1 and 5 voices in Layer 2. If this already gives up to 50% on a Quad Core system i guess that the unison has to be optimized or that the high unison settings are exclusively for Quad Core CPUs.

If i set both layers of that patch ("PAD coldsmoother III SM") to 4 voices unison i get 26% CPU on a Q6600 CPU, 2 voices unison give 16%, completely without to 10%. All results if a chord (3 notes) is played.

[kv331]

UPDATE: I found that the high CPU use in rsmus7's pad sounds is caused by high unison settings. The "PAD coldsmoother III SM" patch uses 8 voices unison in Layer 1 and 5 voices in Layer 2. If this already gives up to 50% on a Quad Core system i guess that the unison has to be optimized or that the high unison settings are exclusively for Quad Core CPUs.

If i set both layers of that patch ("PAD coldsmoother III SM") to 4 voices unison i get 26% CPU on a Q6600 CPU, 2 voices unison give 16%, completely without to 10%. All results if a chord (3 notes) is played.

I'll try to figure out the problem with those pads.

[kv331]

I'll try to figure out the problem with those pads.

Apparently he used 13 voices unison on 2 layers plus all sorts of modulations

[Lotuzia]

Save VS Save As Bug :

When you save a preset in a particular folder with Save As ( Like in this LeLotusBleu folder ) and then make edits on the preset and use only the Save Button : The preset is saved at the root of the preset folder (and NOT in the right Sub Folder it comes from) meaning you end with a bunch of duplicate presets. ( I had around 20 )

Anyhow : Please Bulent we still need a VISUAL indication when we use the Save function. ( Though obviously I wont use it for a long time now, until the bug is fixed )

Nothing's worse than saving an instrument after a long session, then finding oneself think " Ooopps did something happen ? Is my preset saved or not ? And in the right location ? ....."

Then thank you for considering enhancing the reverb, and the delay still doesnt preserve any stereo settings made on oscillator/layer stage when not set to "ping pong mode". Meaning that the signal is probably summed in mono before the delay if its not in ping pong mode. If this situation should stay the same, it might be a good option to make the delay default as ping pong to preserve the integrity of the stereo scene. Well you'll see.

Any enhancements of the Mseg would be very KOooOl AND usefull AND powerfull. Thanks ofr that again.

Cheers,

L.

[kv331]

Meaning that the signal is probably summed in mono before the delay if its not in ping pong mode.

Hey, that's weird. Just tested and yes, the stereo output is mixed with the left input only

Last night Brian reported the same issue as well.

EDIT: Just fixed it.

[Ingonator]

I'll try to figure out the problem with those pads.

Apparently he used 13 voices unison on 2 layers plus all sorts of modulations

That's what i already said in the update of my last post. The result is that a patch with two times 8 voice unison could almost bring a Q6600 CPU to it's limit.
Based on that my wish to the patch designers is to use very high settings only when it's necessary.
My experience is that the additive oscillators when used as a "supersaws" mostly deliver better results than with using unison resulting in much less CPU use.

Many people complained about the CPU use of Largo and Synth Squad so Synthmaster 2 should not do the same "failure".

[kv331]

Many people complained about the CPU use of Largo and Synth Squad so Synthmaster 2 should not do the same "failure".

Dont worry, I'll do my best for that not to happen, even if it requires editing patches one by one to optimize them

[kv331]

Save VS Save As Bug :

When you save a preset in a particular folder with Save As ( Like in this LeLotusBleu folder ) and then make edits on the preset and use only the Save Button : The preset is saved at the root of the preset folder (and NOT in the right Sub Folder it comes from) meaning you end with a bunch of duplicate presets. ( I had around 20 )

L.

I thought I fixed that one a couple of build ago, I'll double check now.

EDIT: This one was already fixed on build v1.9.9.19: FIX: When saving existing presets, the preset is saved to the root folder if it existed under a subfolder before.

[Lotuzia]

Save VS Save As Bug :

When you save a preset in a particular folder with Save As ( Like in this LeLotusBleu folder ) and then make edits on the preset and use only the Save Button : The preset is saved at the root of the preset folder (and NOT in the right Sub Folder it comes from) meaning you end with a bunch of duplicate presets. ( I had around 20 )

L.

I thought I fixed that one a couple of build ago, I'll double check now.

EDIT: This one was already fixed on build v1.9.9.19: FIX: When saving existing presets, the preset is saved to the root folder if it existed under a subfolder before.

Oh yes My bad bad Bulent I'm still in the previous build Sorry for this one

Btw Bulent, could you please check if a kind of similar stereo bug to the delay one does not affect the reverb itself in its current state, cos the reverb also does not preserve the integrity of the stereo scene. ( But here it might be inherent to the particular reverb you coded with the Stereo parameters aso)

As for the delay I first reported this bug well ..... one month ago at least But I guess it disappeared in the "to do" list. SM2 is growing fast, its probably difficult to keep track of everything

Another concern is that I cant find a way to exchange the old and new Default Skins for the presets I made before the new Default skin was made. Is there a way to do that or are all the old presets condamned to use the old default skin ?

Cheers,

L.

[Cyforce]

i getting crazy....

before 10 minutes, i worked on sm2 as vst in logic 5.5.1 as host. build a bass, write the present info, and click to save as...

BAM - short freeze, bluescreen, laptop reboot


before 5 minutes, open i sm2 as standalone... change osc to wave, seletected a wave, pitch down, go to osc2, chang to wave, open a wave, open another wave...

BAM - sm2 quit and show me the error windows


How can i build patches if sm2 crashed on every use atm?
as vst it crash if i open some waves, sometimes he don't crash if i open waves, but than he crash if i want to save the present....

[kv331]

Did you get the latest build? (21)

[Cyforce]

Yes!

[S-N-S]

Did you get the latest build? (21)

didnt know build 21 was released yet??

[Ingonator]

My feeling is that at some point now no new features should be added (i can't really find really important additions at the moment...) and instead the focus should be on doing the rest of the presets and improving both stability and performance. Looks like with every build some new problems appear or some old problems seem to come back.

[kv331]

This crash has to do with delay loading wavetables to decrease initial startup time of sm2