I’m trying to use some of the basic DSP operations, such as the parametric EQ just to emulate “LOUDNESS” when playing music quietly.
However, any time I adjust a parameter, the audio stops for several seconds and so it becomes really difficult to try and tweak the sound.
I’m used to being able to do this kind of thing trivially in real-time with such things as plugins in DAWs that are playing multiple tracks of audio, even at very high sample rates and low buffer sizes and essentially in real time.
However, I don’t know whether this is a function of the server I’m using (a recent vintage Synology NAS) or inherent in the software. If the former, I’m happy to upgrade to whatever is recommended to allow this kind of thing to be done in real time but I’m a little surprised this doesn’t work.
I should be able to do things like adjust the frequency and Gain in each band in real-time.
I too have been spoiled by the real-time parameter changes within a DAW.
Would be really cool if this could be implemented in to roon. Not sure what the difference is between the way roon handles DSP and a DAW handles DSP.
I’ve had the ability to tweak settings in real time in a DAW for 15 years or so, so I can’t imagine it’s anything to do with the horsepower required to make it happen.
Maybe roon operates with some inherent latency that can not be removed?
Exactly – you can have hundreds of tracks (each one running at 48k or even 96k) and still tweak EQ **i(often with some very sophisticated EQ algorthms, not just simple parametric or multi-band filters) in real time on both the individual tracks as well as on the master out and hear the results in real-time.
While I’d rather not, I’d settle for a slight delay before you actually hear the change but having the actually audio drop out for several seconds when you just change the amount of gain is absurd.
I wanted to share some information from our CTO Brian, albeit from quite some time ago, that is still relevant and applicable to this discussion. I’ve copied the relevant part of it below:
There’s about 2-30 seconds of buffer between the DSP engine and the output device, so to apply settings we actually need to rewind and re process that audio, then begin playback again. This is all because we are architected around streaming and distribution, so everything is capable of supporting really large buffers. That is why we don’t do.any real-time filter manipulation–the lag would be a worse experience than the hiccup.
Why can’t you simply apply the DSP processing to “new” data such that at some point (in that 2-30 seconds range) the effect will kick in?
In my opinion, if you’re doing things like EQ tweaking, having the effect kick in (and so slightly changing the “tone”) a few seconds later is going to be far less disruptive than actually stopping the music. You’ll suddenly hear a little more bass, for example …
My take is that it wouldn’t be instantaneously reflected and unless you are doing micro level tweaking and waiting for the small change to be reflected, you wont know if it was something you changed 1,3,5,9,10,15 seconds or more ago
Well, it wouldn’t be instantaneous (because of the buffer) but I would still prefer a delayed change over a complete interruption in the sound
Well, I was wanting to do small tweaking - e.g., change the bass gain by 0.5db, for example, to implement “Loudness” at soft volumes, for example
That said, with respect, these kinds of systems (Roon, connected to high end streamers, connected to high end amps and speakers) are intended for audiophiles who supposedly can hear the minute differences between CD and FLAC o (at the same bit level and sampling rate, if you’re to believe all the claims made in forums all over the place) so surely such listeners would be easily able to tell the difference.
It has been said that the human ear/brain needs 3db of change to determine a difference IIRC…don’t quote me I might have got that wrong but then Audiophiles have some interesting talents indeed.
Well, I can’t speak for everyone (and not even myself) but I will tell you that the record producer who mixes our albums has an astonishing ability to hear absurdly tiny changes…he’ll often tell me to increase my volume by 1dB and can tell (by ear) when I’ve done it.
That said, if the minimum change is 3dB, fine — I still don’t want the music to be interrupted when I make a change!
