Cubase VST and virtual drum samplers....

[captain54]

I've read some recent posts on users and their experiences with vdrums and Gigasampler/Studio, but I was wondering if anyone has any success in playing/recording in realtime with an edrum/softsampler/DAW setup...

I using Yamaha dtxpress triggers, a Hart Accusnare, and Cubase vst32/5.0, with the LM-4, Ao DKI, and have also used Unity....

With a Midman Delta 44, the ASIO drivers bring a latency of about 9ms, but its shame you can still hear a slight delay....

Its very cool to be able to play your sampled Yamaha Stage Custom Kit, with killer realistic cymbals.....however, its hard to really record in real time with even this slight delay...

Has anyone attempted anything similiar to what I'm doing with any success...??

[e-drummer]

Hi Captain.
I use VST32 as well and at the moment there is no way round it.
If you are happy to record the midi output from your kit listening to the Yamaha sounds and then just use the LM4 on play back then its not a problem, as you well know the LM4 is internal to VST so there is no latency on playback.
I feel however that what we are hearing dictates the way we play as in how hard we hit the drums with individual limbs in relation to each other.I guess you could say that we as drummers are "mixing" the overall drum sound as we play.
I`m pretty sure that all the big software and hardware companies are working to try and get this latency down to nothing.
Ive also heard from people who have had a latency of 6ms with giga and they say they can still hear a difference so I guess until were at zero latency were are not going to be able to record in realtime whilst actually listening to the sounds that are in computer.

I am beginning to think the weakest link in the set up is actually the drum trigger module itself in that it has to convert the incoming pad signal to midi and then send it out to the pc which then has to map it to the relevent software sound.
If we had a dedicated software module that actually handled the signal from the pads as well then we may get nearer to zero latency.

Sorry I haven`t really answered your question but it seems to me that this is being held back by the technology at the moment.

[captain54]

Thanks e-drummer....

another poster from a Cubase forum using the Alesis Dm4, vst32, and the rm2 from LinPlug claims he's managed 4ms with all the buffers set to their minimum..this means minimum plugins and audio tracks while recording...

The rm2 seems to be the most responsive to a real time recording sesssion, with AO dk1 2nd, and the Lm-4 lagging behind....

You bring up an interesting point about the lack of sophistication in most current midi modules for this application, but there also is obviously a difference between softsamplers...

Yes, we probably won't be satisfied until we can achieve zero latency, however, I've experimented in a live stage situation, and if you are monitoring yourself live with a v-drum setup, there is 3ms latency from your ears picking up your performance thru stage monitors...

If you are playing an acoustic set live, the
same applies if you are miked and hearing yourself through stage monitors....

Furthermore, 3ms is about the delay you would hear from your guitar players amp to your ears...

But like you said, it depends to on your playing style...Bottom line, the current technology, with its flaws, opens up world of creative possibilites, I just hope the technology can keep up with our imaginations...

[Ace]

Ahh, the old 'delay' chestnut. Time for a little reality bite.

First, don't worry about the latency of your soundcard settings. In fact, don't worry about using your soundcard at all to start off with. The reason that everyone has a worry about ASIO and latency is more to do with the fact that it causes problems for monitoring. The solution is not to monitor from your computer. You have some great gear there - use it!

Get a sound from your drum gear that you are happy to play with and play along to the music without monitoring from the computer. If you listen to the output of your modules then you don't have to wait for the data to go into the machine and then be spat out again. Let the midi side simply act as a recorder. Then when you listen back, get the unit to play back your performance on your existing drum gear.

If you want to talk about delays, you should know that there isn't a drum to midi converter around that can get the signal down the midi lead in under 9ms as midi has inherent speed limitations.

Once you have a good PERFORMANCE, then tweek the sounds. Remember, you can have the greatest sound generating gear in the world and still sound lousy if the performance is no good. Concentrate on sound design after the fact - that is the big benefit of computer-based systems : you leave your options open.

Most of all, stuff the technical side of it - HAVE FUN! Remember, this is music, not quantum physics....

Dave

------------------
Simmons SDX (x3), Roland TD7, Alesis DM5, Simmons pads, Hart e-cymbals,

[This message has been edited by Ace (edited January 17, 2001).]

[captain54]

Thank you Dave for the reality bite....

(Okay, reality, you can let go now)

Your method (applying VST instrumentation after the fact) is really the only way to go
right now without completely pulling my hair out of my head...

To me, playing actual Yamaha acoustic samples triggered from e-drum pads is a total rush....I can even load samples of Ringo's old Ludwig kit (not the actual kit of course), and to me this is totally astounding to be able to do this in real time...

Your totally on in your statement that its all about the performance...and while its fun to bang away, a tight performance is no where near possible monitoring thru the vst instrument...

thanks for your words of wisdom...

[sox]

I use a TD-8 kit with VST 5 and record the data as MIDI. Send it back and the TD-8 responds exactly as I've played. Depending on what I want to do with the material, I'll either rerecord the whole performance as audio in VST or sample chunks of it in a hardware sampler.
You could take audio samples of 3 different levels from your kit (per drum)and also load them into one of the VST software drum machines (quite a few of the new ones I'm seeing lately are multi-sample capable responding by velocity) if you wanted to go this route.
Sounds like a lot of work to me though.....

[captain54]

A lot of screwing around, I agree, but not a bad suggestion though...

For what's its worth, I was messing around in another studio and we got the Logic Platinum 4.6 and its integrated sampler, the EXS 24 working with triggers and the timing was almost dead on ..I couldnt tell that much difference between that and a standard midi module as far as timing goes...

Maybe its because its not VST....

Still in a real time recording session, I would be a little uneasy relying on this method to record...

[juggernaut]

Working with TD-6, VST 5.1 and triggering Proppelerhead Reasons drum module through REWIRE. Working almost flawlessly. Took me about 6 months of experimentation to finally find the right sound card, the right software and the right configs.

First of all, with a Delta 1010 you can get your latency down to 5 ms (which is not noticeable) at 44.1 khz and buffer settings of 256k I believe. You can get it even lower than this, but the audio may start choking, especially if your using effects. I find these settings a good happy medium.

Here is the kicker: Most people think latency is the problem, but really it is your "audio to midi delay" settings inside of cubase. The TD-6 requires a setting of about 95 ms. THis offsets the midi so that when you record and playback your recording, the 2 sync properly. God it took me forever to realize this!

Now I can monitor my playing through cubase with effects on the samples I have loaded into Reason's drum module (ReDrum). Only thing I havent figured out yet is how to get the hi-hat pedal working properly. I'm looking for a workaround on this. I will let ya'll know when I do.

If you have any questions about running a set-up like this, give me a reply. Thanks!


[This message has been edited by juggernaut (edited June 10, 2002).]

[This message has been edited by juggernaut (edited June 17, 2002).]