I seem to recall you (Steve R...) asking some kind of question about digital filters on sweetwater or mix forums, but i'm not sure if it was you or not. At any rate, Somebody was asking about the differences between software filters and resolutions. This question has been on my mind for some time. Actually, at one point I emailed the makers of DDCLip to check with them to find out how good the interpolation filters are in their 96kHz program. It was an interesting response I got, because I don't fully understand the nature of interpolation algorithms and such. I can say that they use high-quality filters and attempt to reduce ripple. It seemed to me that they are aware of these kinds of issues.

Personally, from using various fft and eq programs, I have noticed a correlation between sample/frame/bin/window size and bass response. Bigger unit sizes usually mean more bass, but slower processing time. Smaller units usually mean less bass, but faster processing time, and perhaps some high-frequency boosting, due to aliasing errors or more computational power available.

It seems that 256unit functions are routinely fastest, while 8192 or 16384unit functions have less spectral errors and frequency-response distortions, but at the price of much more memory needed, and perhaps a sacrifice in quality if real-time power is needed and the system runs out of processing power.
<BGSOUND SRC="http://ajsblue.virtualave.net/wav/dragnet.wav">
I think that some programs actually skip frames/bins/samples/windows when they get low on memory. This would imply that bass response of particular plug-in or algorithm or program could fluctuate according to the CPU time resources and RAM resources available during use. It would suggest that better results in audio fidelity would be achieved by always rendering multitrack compositions to disk in non-real-time, so that the computer can get it's computational power and resources as it needs to.

It would also imply that certain plug-ins are susceptable to frequency response distortions according to memory conditions. However, some programs might compensate for this balance of computational power issue by sacrificing bit resolution in order to save sampling resolution. This is of course something that happens in mp3 and minidisc (ATRAC) compression filters. I strongly intuit that this happens. It would ensure that the stream of audio data is rather constant, despite less resources to go around for all of the plug-ins. Since rendering processes can take advantage of virtual memory, while many real-time (live) processes can't, I suspect that non-realtime rendering of audio files typically produces slightly more reliable results under some circumstances.

In my experiences, general audio program performance increases when Virtual Memory is disable and the cache file deleted or reduced. With this new idea in hand, however, now I'm seeing how VM could be useful in non-realtime audio programs. In general, however, I've found that VM is a bit more stable on Macs than on PC's, but that is all rapidly changing with the releases of new operating system designs. It is also more complex because of the use of RAM for disk caches and video caches and such.

Still, considering the computer as a battlefield over RAM and CPU power, it makes sense that more RAM would yield better filters since many digital filters rely heavily upon digital delays, and digital delays rely directly upon RAM amounts.

Since VM is slow, it's probably used less for delay/eq calculations since it's delay time is not dependent upon audio, but upon system circumstances. Most likely, some programs deal with this by handling the memory needed for delays directly, and not relying too much upon calls to the system for more memory, since the system may be late in returning the free memory. Again, another possible distinction between EQ/delay algorithms/programs/plug-ins. Most likely, the more predictable systems have worked out a very speedy and no-questions way of accessing and budgeting RAM. I imagine that the VST and DirectX and TDM protocols standardize how this happens. Maybe each has strengths and weaknesses that can be exploited by those who make the plug-ins that go with them.
<b>
Another issues which I have considered in this mundane topic is that supposing an audio plug-in runs out of RAM or CPU time, but can't afford to wait for it to come. I'll bet that some quick interpolation or random noise or faking it happens just to keep the chain flowing and keep the plug-in train from derailing over a few lost samples. Maybe hundreds or thousands of samples can be lost but smoothed out, or replaced with noise. We may never know! Considering that plug-ins are all about munging up the audio, anything is possible.

OK. And now for an even funnier look at software mixers. In digital mixers, it's possible to mix two separate channels just by switching back and forth between the two channels at a rate that is fast enough to be unnoticed. I know this, because I have experimented with this kind of splicing or dithering.
Basically, the new waveform is just 1/2 of each of the original waves, but in alternating stripes. When done correctly, this will yield something that sounds like a mixture of the two sounds, but without any amplitude clipping, since the audio data is multiplexed and not added.

I suspect that this is going on with some programs and even hardware digital mixers, because it's still a tricky computational procedure to mathematically add(mix) 2 or more 24-32bit samples, especially when there are gazillions of such samples in a continuous stream. Probably many standalone digital mixers cheat by actually mixing the data in an internal analog mixer. Heck, maybe even many soundcards do it. It would imply that internal digital to analog to digital conversions are happening. Another way that some products might SUCK, while others become RAD. Mixing in the analog domain has the advantage of possibly more headroom, and thus less clipping. But it also could mean more analog noise introduced, and for sure, quantization noise. Still, some manufacturers could employ quick analog equalization during the mix, so maybe noise could reduced but frequency response changed.

You see, it all depends. When it comes down to it, we can't know for sure what's going on in digital hardware. We can't even know for sure if all of the digital hardware is digital! There are digitally controlled amplifiers and digitally controlled eqs, and there are digital amplifiers and digital eqs. There is a difference. The digitally controlled devices are merely analog devices "with robots controlling them". I think it's time for a break from this rant. I hope that I have planted the seeds of questions in someone's mind out there in this web-wide-world.

<A HREF="http://arts.ucsc.edu/EMS/Music/tech_background/TE-09/teces_09.html">This is rather interesting. It is a nice site that explains in simple pictures and text how various things work. This page is about digital delay and digital EQ. </A>

<A HREF="http://www.google.com/search?num=100&hl=en&as_qdr=all&q=how+to+build+an+analog+to+digital+converter">And for the really really ambitious, here are links for building your own Analog to Digital converters and messing with the PC serial and parallel ports!</A>

Hi, I'm writing a converter. So I need the format for .dss files. Is it possible to get the format? Please help.

Hi,
Can anyone tell me if there is any software out there that will remove bass lines from songs please?
Or at least if not removing it, reducing it so that it is very quiet.
Cheers.

I am trying to find out what machine to buy to be able to play music videos on TV with words on screen, like Karoke night at night clubs. Can you help me with the name of machine and how to buy the video to go with them...

I have tried three times this morning to download the mp7.1 media player and all 3 attempts were unsucessful.
It appeared to be loading for over an hour each time, but each time aft erward when I tried pulling it up I was not able to locate any of the downloads. This may be the fault of equipment or personnel and realizing those possibilities I almost did not write to you today . However, if you do have any suggestions, questions, and/or comments please let me know. I appreciate any advise and/or information obtainable from you. Thank You So Very Much For Your Time.......
Ralph Mancuso 3119 W. Cochise Dr. 115, Phx, Az 85051
602 678-6964 rgm115@aol.com

Can someone suggest a software program that can help me
create electronic, trance music ?

I need OPTI931 driver for win me