Wish List

I do not know, whether it is possible, but it would be great to be able to store some audio data in built in memory and than have a module which allows to play it from there. Like loading a short audio file from computer into the DSP's memory and then play it on its own, or using as some inner source for subsequent processing and so on.

Hi Skfir,

1. You don't need to terminate unused outputs. For unused inputs, I would suggest just using a DC Source (from the sources folder) and setting its value to zero.

2. I like this idea. I will forward it to the software team.

3. Unfortunately the limited data RAM on the SigmaDSP makes it impractical to store audio and play it back from the DSP. Even on our SigmaDSP with the largest memory out of the entire product family, the audio clip would be limited to a few hundred milliseconds, and it would occupy the RAM so that you effectively wouldn't be able to do other processing.

Thanks!

Sorry, all my wishes will mainly be about oscillators, since it is what I use the most. Along with a linear data controlled sweep it would be great if current VCO oscillator had sort of a check box, which would let it oscillate until it meets the first zero crossing, even if it is zero on the control pin. The idea is now it drops oscillate immediately after it senses zero on the control pin. So if making a sweep oscillator, with triangular wave as a driver, it may stop in the middle of a cycle, which causes a terrible click, which in turn on high frequencies is almost impossible to eliminate even with help of sweep mux, since there is simply not enough time. So zero crossing on VCO would me amazing.

It would be extremely useful, and I believe not only I think so, so have an ability in square wave oscillator to adjust duty cycle. It would be very very useful.

Yes, a "voltage" controlled PWM would be handy. We can make one now with the triangle-wave gen and comparator, but having it in one block is more convenient.

Reverb! Would that be possible? Is there enough memory for this effect? Thanks!

SigmaStudio places Reverb in the toolbox for the AD1940 or ADAU144x DSPs as these have the most memory (the ADAU1442 includes 8k words of data RAM). Please see:

https://ez.analog.com/message/4327#4327

https://ez.analog.com/message/18899#18899

With others such as the ADAU1701, you could feed a delay back into itself with a gain somewhat below unity. This achieves a lo-fi reverb similar to that often found in karaoke machines. Now that I've tried it, no -- it's not even good enough for that. There's also enough RAM for chorus and flange effects, but not for a "concert hall" reverb. So, what we wish for is more silicon.

Bob

Message was edited by: Robert Blowsky

Got it! Thanks Robert.

It would also be great to have a voltage controlled sweep, which not only goes from low freq to high, but also from high to low. Because now the start frequency can only be lower than the final one.

And still guys, you may toss rocks at me, but I am very convinced that an ability to play some prerecorded sound from internal memory, even if it is a just several milliseconds may be very useful. Of course I understand that the primary application of sigma DSP is sound processing, but its parameters allow it to be used in much broader way, including various ultrasonic applications. Quite often, in ultrasonic lab things it is necessary to produce very short, but rather complicated sound pulse, (which may consist of non linear sweeps, gaps, differences in amplitude between harmonics, some modelled noises, and so on) and then catch and analyse the response of the system. For example, testing acoustic properties of a musical instrument, or many other things. These complicated impulses are very easy to be made in a PC based applications, like adobe audition etc. by using cutting, pasting, filtering, zero crossing, and some other non-realtime stuff. And most of the times, it is impossible to create those complicated impulses in Sigma studio, or, rarely it is possible, with many limitations, but requires tremendous ingenuity and is still very difficult to tune properly and is very rigid to modifications. But the most important thing is, that it is still impossible to create some really complicated impulse in SigmaStudio and at the same time it is very easy to model it in almost any PC based audio editor. So it would be extremely useful to be able, to make some complicated impulse on a PC, just a couple milliseconds, load it into the DSP, play it when it is needed (by, say, a positive impulse on a control pin), and then do whatever is need with the response. I mean, that "voltage controlled" built in player module still would be of extremal usefulness for many applications primarily ultrasonic ones. Even in spite of the fact, that it is only several millisecond, since in those applications it is more than enough.

Hi Skfir,

Thanks for the wish list. Actually, we're considering adding the ability to play back audio files from RAM in a kind of wavetable synth fashion, as this could be useful in some applications like artificial engine noise generation for electric vehicles. So, you may see this in SigmaStudio some day.