I like the approach of recording the samples themselves as "hot" as possible, and adjusting the gain after for velocity layers. The engineer in me just feels that'll give you the best S/N ratio. In addition, if you normalize the softer samples in youe example above, you'll bring up the noise floor with it. No sense in wasting the headroom.

But the musician in me fears that samples of instruments that vary in timbre [see reply to DA above] won't be recorded accurately, so we're back to a case-by-case assessment. I'd probably go with the first method 90% of the time.

What's next? I don't know ... I'm not quite ready to leave the DS864, as yet. I'm always finding new things. Like in the LFOs: Most LFOs in a mod matrix act in a similar fashion as those in the DS, but these just sound different to me in their results. The three are "wired in series", so to speak, and they modulate each other when set to the same destination. You can get some really convoluted "waveforms" by setting all three to Pitch, Amp, or Pan with different Shapes and key trigger modes.

So what's really next? Tough call. Cyclone is extremely deep in its applications, but nPulse is really a neglected puppy that just needs a little love. PSYN can always be mined for some sparkling gems, but ... final answer? The Spectral Transformer is my single favorite part of the whole Project5 package, and that's saying a lot.

I call it the Black Hole plugin at the studio, and everyone knows exactly what I'm referring to. If you get even remotely close to it, you're sucked in, and nothing escapes. If you're lucky, you pop out the other side into an alternate universe. I'm afraid that once we all dive into that plugin, we may never come back ...

I like the approach of recording the samples themselves as "hot" as possible, and adjusting the gain after for velocity layers. The engineer in me just feels that'll give you the best S/N ratio. In addition, if you normalize the softer samples in youe example above, you'll bring up the noise floor with it. No sense in wasting the headroom.

I am only butting in out of curiosity. when recording the samples as "hot" possible you do not mean to play the sampled instrument w/the same intensity do you?

thanks

have a good one
3Eo

b rock through further spectral analysis i have found that pitch shift that preserves duration has more artifacts in the up pitch shift but pitch shift without preserving duration and just downsampling or upsampling there are noticeably less artifacts either way up or down. i could see no difference but i did not look that close will later tomorrow

edit... oops forgot to say you were right. :-)

I did a really bad thing the other night...

I spent ages putting in different samples into a patch I already created with different sounds of cars rain thunder etc etc etc took me a whole night to get it right, then put in the notes so all together it made a nice listening story inrto for a Hip Hop style track and I didnt save the Flippin patch did I!
and I was deserted with its past state...

OOOOO it rattled my nervs. I felt so sick!



People are mentioning different sampler editors (AKAI) is there a free one out there? That i can build my samples on and then re-open in DS864?

Cheers

C ya

ORIGINAL: PaintedBlue

Here's another question to consider: what are the best practices for managing input levels when recording an instrument at different velocities? In particular, I am wondering how to deal with recording quieter samples for softer-played notes.

For example, I can try to adjust my input level for different velocities to keep all recorded samples at roughly the same level, and I would then adjust the gain for each velocity layer within DS864.

Or, I can record my samples at their natural levels, but I think that the quieter samples would hold less audio data (i.e. and be of lesser quality than if they were recorded at a higher input level?). From there, I could then choose to normalize my samples and manage the gain within the sampler, or I could load the samples into DS864 as recorded.

Any ideas on this?

If you check out the ds864 bank I put on project5.com, you'll see that I used 3 velocity layers. The lower velocites were not normalized. I believe the effect of this is that the final output from ds864 will have the quieter notes even quieter than what was sampled. However, I did this on purpose for that particular bank/program. Other programs I've created I've normalized all the samples (all velocities) and depend on ds864 for creating the final volume level.

Some other thoughts on DS864...

I submitted a few feature requests on the support web form for DS864. I truly hope Cakewalk considers them. I'll share the ideas with you now. Basically, the idea is that because of the cheap selling price of Project5, Cakewalk could sell many copies of P5 simply by touting a great sampler. For example, I know several people who bought Reason (which is much more expensive) just because they like it's sampler so much. P5 has a great start with DS864. The interface concept works well and is dead easy to use. The filter sounds very good imo, and the synth section is quite good. The downfalls are few, but reasonably significant. Firstly, the gui is too small especially for program editing. Secondly, some looping and sample trimming is needed (ie more sample editing). Thirdly, ds864 aliases badly. Perhaps some oversampling and other tricks, along with a quality setting, could be incorporated. At that point you have an easy to use, great sounding sampler that could become a major selling point of P5.

On other matters, I've requested that the Extreme Sample Convertor supports DS864 formats (I've contacted both Cakewalk and the developer of ESC).

sluggo

Ignore that last post, Vienna needs there to be soundfont compatible soundcard attached to the system or else it won't work, thats a shame.

with an pc overhead projector it's not tiny at all. geez i feel kinda bad about complaining now. live and learn. <g>

sluggo do you pitch up or pitch down when you notice the artifacts. theoretically you should see less going down assuming the ds864 is not transposing via DFT function and is utilizing down and up sampling to transpose. also at what interval of transposition do you notice artifacts. they are readily noticeable say if you just import one wav and stretch it to include several octaves (i luv doing this by the way makes some cool sounds) but generally as b rock pointed out you do not want to do this if you are trying to create an authentic sounding sampled instrument. i plan to run spectral analysis specific to the ds864 as soon as i can.

after all, DS864 actually IS a sample editor

DA, I don't think anyone wants it to include full-fledged sample editing a la Sound Forge, etc., but neither do we want a kiddie editor coded by Mattel. I think that we're looking for an intermediate solution that allows the user to go beyond the bare basics, and some easier (and visible) methods of constructing a complex sample set.

Even a toggled [full-size of the GUI] window would be a vast improvement, but I'd like to a better way of sample import, a method of locking samples and velocity layers out of the editing process [beyond being unselected; too flaky], root note, loop point, and adjustable span choices that can be selected and saved as part of the import dialogue, etc. That's just the basics, and off the top of my head; the real wish list is a bit larger.

There's quite a bit of bugginess that needs to be addressed, as well, including a huge bug when opening the DS864 in Sonar, using the Insert Synth method and automation. As another example, try to select the Free switch in an EG on the first go-'round; can't be done!

Personally, I don't find the interface easy to work with at all for what I'd like to accomplish, and that is some complex sample sets that highlight the excellent synth section processing the samples. I was hoping that Cake might have a comment on the topic, considering the depth of discussion here, but I understand the reason that they haven't: they can't see the thread. Can you imagine how long it would take for your vision to recover if you coded this sampler?

I just have to disagree on that point I guess, B-rock.
I dont want anything MORE hi-tech and complex than what I currently have with DS864...mind you, I understand the bugs and the size issues with it.
But I think in keeping with the scope of the program it was perfectly krafted for guys like me.

Understandibly, you "tweakerz" need more bells and whistles....so, thank GOD for the marketing department!

Digital Aura you must admit banks and patches for the ds864 have not been forth coming. this i believe has led to a desire on a large number of p5 users to want to build there own. ds864 size alone is one big pain in the arse. arse is that a canadian or uk word i have no idea where it came from. never before out of my mouth or typed from my fingers. i don't believe making it better will make it more difficult to use.

anybody know if it's illegal to sample an instrument that uses samples? like the digital synth of today

have a good one
3Eo

ORIGINAL: three_eyed_otter

anybody know if it's illegal to sample an instrument that uses samples? like the digital synth of today

I remember I read an article in the german keyboards magazine. It says that you can sample these instruments because they are not copyright-protected. If you sample from korg, don't use Korg in the sound name. I'm not sure ybout this, the answer is surely complex.

i not a lawyer but would be willing to bet if you sampled a synth that uses samples that are copyrighted you would be in violation because essentially you would be just transfering from one medium to another. copyright laws vary from country to country. in the eu 50 years after a work is recorded it may be sampled in the usa it is 70 years after the death of the author. so copyright laws vary and you are required to be within the legalities of the country or countries you are selling your art in. the eu will be able to sample legally early elvis shortly but the usa has at least another 45 years or so to go before it can. you must hire legal representation either way to secure your rights. copyright law is very complicated and there are no clear cut rules. you must always hire a lawyer or pay the consequences.


edit.. i should clarify; the above 1st sentence refers to compiling samples to sell you are certainly free to move from one medium to another in order to make music. you can usually look at your license agreement to see what rights you are entitled to. most license agreements today forbid repackaging of the samples to sell.

in the U.S. citizens pay roughly 50% of their income to the United States of America Government while U.S. corporations pay a paltry 7%.

have a good one
3Eo

does anybody know if it is legal to sample a rompler say like Atmosphere or DKFh 2?

oh i mean using the DS864

I'm looking into how the Akai S5000 format performs its magic on .wav files, making for a compact download.

There's an interesting, and somewhat related, parallel thread developing at the Sonar forum. It asks the question of whether the Soundfont format is a "toy" format, or if it's inherently capable of high-quality sampling. [Read the entire thread here.]

In the course of the discussion, it's been stated that .sf2 files are limited to 16 bits (evidently not a problem for the DS864, with a simliar ceiling]. It also postulates that the Akai format is highly compressed, as it was originally developed for hardware with very limited memory capacities. That would make a lot of sense. I didn't notice that the Hollow Sun samples suffered as a result, but I'm sure that I'm subconsciously compensating for that fact by considering the age & condition of the original hardware that's sampled.

One route to take that was brought up was using CDXtract as a conversion device. I've barely skimmed over the home page for that, but it looks like that may hold some promise for solving our own tribulations here. I suppose that I could also contact Rubber Chicken to see if there's any hope for a DS864 conversion process in the near future. but I admire the process (so far) that I've glanced at with CDXtract. I'd like to minimize any conversion errors to the sounds themselves, due to filter differences, etc. I know that we'll lose any synth module settings in converting formats. As the Sonar thread also mentioned, it's best to utilize a sampler's native format; anything else is a conversion, and as such, subject to or even prone to error.

Edit: I corrected for the mislabeled product name. CDXtract isn't cheap for what it is (a utility with a narrow user base), but it does have some great tools for batch processing and velocity stacking that might help justify the price (for me, anyway).