Stereo Tool

Building now.

Upsampling latency for processing sample rate back to input sample rate (44.1 kHz --> 176.4 kHz or 48 kHz --> 192 kHz) was 128 samples (512 for latency 1024), now 0.

Downsampling latency is still there (but I think I can also reduce it to very close to 0). If your input is 44.1 kHz or 48 kHz, that's no issue.

FM upsampling still causes some latency also -> will be fixed.

Composite limiter also causes a bit of latency -> that will be more tricky to fix (and I'll have to check how much it is, if it's just 1 ms it might not be worth the effort). --> Edit: I think it's only 0.7 ms; 2.8 ms for latency 4096. I'll measure it later to be sure, but if it's really that low it's acceptable for now.

(The issue here is: Non-FM output has exactly the latency as specified, eg. 512, 1024, 2048 or 4096 samples. FM output has some extra latency. Which is bad. And it can be solved.)


One thing does cause a much bigger latency: FM Transmitter calibration. 512 samples at low latencies, 2048 at latency 4096. If I want to fix that, I'll probably also have to fix the 0.7 ms composite limiter latency...

- Issue with manually changed multiband frequencies fixed
- Slightly lower latency if input sample rate is above 60 kHz.

BETA601-005:
Winamp DSP plugin: http://www.stereotool.com/download/dsp_ ... 01-005.exe
Stand alone version: http://www.stereotool.com/download/ster ... 01-005.exe
VST version: http://www.stereotool.com/download/vst_ ... 01-005.dll
VST version (No SSE2): http://www.stereotool.com/download/vst_ ... 01-005.dll
Command line version: http://www.stereotool.com/download/ster ... 01-005.exe
Linux command line version: http://www.stereotool.com/download/ster ... ETA601-005 [not available]
Linux GUI version: http://www.stereotool.com/download/ster ... ETA601-005 [not available]

Thanks for the quick response Hans, but there is still an issue with the 4096 latency setting. The whole audio sounds like it is being chopped by a low frequency square wave.

_________________
J0HN II

HanzZZZZZZZZZZzzzzzz you are the best!!!!

BUT, i ve got 2 simple questions that matters )

1 ) is " bass boost " ....BEFORE the " bandpass filter " ??

2 ) how does the " dirty bass " work ? can i use it for FM broadcast ?

John: I'll run a test with your settings.

camclone:
1: No, afterwards!
2: It just tries a little less hard to avoid bass distortion. You can definitely use it for FM broadcasts (or anything else). Some people like the effect, others don't.

In regards to his question, and partially to do with mine that's unanswered about the clipping in multiband:

Does the "Maximum boost from" slider then take into account the setting in bandpass and shift the boost start point up to the highpass setting? In other words, I have a highpass at 35 Hz, which is supposedly the bottom-end of my Logitech Z-2200s. If I leave the Maximum boost slider at 0, is the max boost at 0 or 35?

FM Output latency. That's important! And there is one thing there. Standalone.
At buffer size i can't set it under 0.02sec using Juli@ and Asio. 0.02sec mean 20ms. Is it really 20ms? I know that asio can go almost to 2ms (4ms safe).
If i try to set under 0.02sec stuttering occures. Just checked again at transmitter site 0.017sec is minimal what i can set, currently it is at 0.018.

With what latency setting is this?

Processing takes some CPU time. If the latency is higher, more data needs to be processed - in bursts.

If this occurs at latency 512 etc., something is really wrong. At latency 4096 I would expect something like this. (Basically, the minimum setting would be 4096 samples / 2 * CPU load, so for a CPU load of 50% (at the core where ST runs), that would be 2048 / 44100 * .5 = 23 ms.

If the latency is 512 samples, the minimum ASIO latency would drop to 256 / 44100 * .5 = 3 ms.

First any audio below 35 Hz is removed (well, theoretically. The filter isn't that good. It does get very strict when you select 'non-phase linear' filtering.

Then it just processes the audio. At that point it doesn't matter whether there was audio below 35 Hz in the input signal or not: It's not there anymore so nothing is done with it.

So audio below 35 Hz 'does not exist' as far as the Bass Boost filter is concerned. And it doesn't matter why it doesn't exist (wasn't there in the first place, or was filtered out).

Hope this answers your question...

Is that new in this new BETA (005)? If so, are you using a higher (> 48 kHz) input sampling rate? I don't fully trust my new upsampling filter yet...