SD latency as dXi plugin

I'm running Sonar and 2-3 soft synths including reaktor, as well as dXi insert effects (including SD).

I noticed that when I add SD as an FX insert it causes the track output to be delayed by an amount that seems to depend on the track. For example, I have a soft synth with a SD insert effect. With SD activated, the track runs about 300 millisec behind, and when I turn it off it's fine again. This is not the case for the other plugin FX I'm using - only SD. And it's not a question of the dry/wet mix: even with the SD mix set to 100% dry, the sound is delayed by some fixed amount. In one case even about half a second.

This means I need to offset the track by the same amount to compensate, which seems silly somehow, since I need to manually adjust the offset until it sound about right.

Is it normal for SD to have this built-in latency? I've already set its ASIO latency to the minimum of 32 samples, but the real latency is much higher. Isn't there a way to synchronize it with the rest without shifting the MIDI track in time?

Suggestions?

cheers, thephaseshift

PS wenn's für Sie einfacher ist, auf Deutsch zu schreiben, geht das auch i.O.!

 

I get the same problem in Logic using SD as a vst insert. I spoke to NI support about it and they could reproduce the problem, but not reliably. Therefore, to me at least, it seems to be a bug. They also said it would probably be a long time before anyone gets a chance to look into fixing it.
I wonder if NI plan more development of SD?

 

Great!!!!
.. I just bought SD for use as an insert FX in Cubase SX. Wish I had of known of this bug before I let go of my hard earned cash ... plus judging by the traffic on here I doubt they are going to waste time/money fixing it.
-grimley

 

There is one nasty workaround for this SD 'latency' problem that I ended up using sucessfully in Logic. I recently purchased Cubase and I haven't tried this workaround with it yet, but if I did, this is how I would do it...

-Add two group channel tracks.
-Insert an instance of SD into both group channels.
-In the first group, switch SD to bypass using the Bypass button within SD itself, NOT the channel's VST insert Bypass button. (with it's bypass switched on, SD will leave the signal passing through it dry, although the SD 'latency' will still be there)
-In the second group, set up SD for the effect you want, 100% wet.
-Route the output bus of all audio chanels to the first group channel.
-On all channels you want to be affected by SD, set up one of their channel Sends to be routed to the second group channel. Use the Send amount to obtain your required dry/wet ratio.

This should ensure all audio signals, SD affected or not, have the same degree of 'SD latency' applied to them.
This is an ugly solution, and a waste of cpu power, but it worked for me. I hope it helps.

It's a pity NI seems to ignore this serious SD problem. It completely undermines SD as a useful tool for any situation where it is required as a plugin effect.

 

Your problem is too LITTLE latency,I think.I presume you're still using your TerraTec-Try setting your ASIO buffer on the card to 256 samples(about 5ms at 44.1).SD likes larger buffers-with really small buffers like you're talking about,there's just not enough time to do the computation.Why 256?Because SD likes buffer sizes in powers of 2,and 128 is too problematic.I use it all the time in Sonar with that buffer size(256) with no problems.
ew