Way to verify buffer is not too low? [SOLVED]

[jasonwagner]

Hello Folks,

 

For initial tracking of bands I record keeper tracks of drums and bass but everyone else is also playing, singing.

 

I am using a hearback system and symphony i/o with the PCI card. Trying to get the lowest latency possible I work with a 32 buffer, which with the syphony and the hearback I have a total of 3.4ms latency, which is very workable.

 

My question, is there a way to verify the computer can handle this ok without creating pops and clicks that I read can come from too low of a buffer. I set up 32 tracks for a worst case scenario and record enabled all of them and recorded for about 4 minutes, I didn't notice anything or hear any clicks.

 

Is that in your opinion a good enough test to say, going forward I shouldn't have any surprises? Is there anything else I can check to make sure the system isn't struggling? The CPU meter and HD meters seemed fine, as in they was no spiking.

 

Thank you, I know nothing is ever 100% for sure but I would like to verify anything I can to ensure stability while tracking with that low of a buffer.

[David Nahmani]

Is that in your opinion a good enough test

Yes!

 

Of course, surprises can always occur, and the ultimate test is to just do it a bunch of times in a real world scenario. But for now that's the best way you have to test it.

[jasonwagner]

Thanks David!

[ski]

The best way to test this, IMO, is to make test recordings with no input signals (record silence on every track). Then go into the sample editor on each recorded region and use the Change Gain function to boost the gain 24 dB. It will take only a minute or two, and the purpose is to amplify the level of any low-level clicks/pops/spikes that might exist in the recording (due to low buffer size). Then, in the arrange page, zoom in sufficiently to fill the entire arrange area with your tracks. Next, click/hold the little waveform icon in the lower RH side of the arrange area and boost the waveform height until you see a clear representation of each track's waveform in the display. Scroll through the arrangement and look for noticeable spikes in the waveform. If you see any, it could be a sign that your buffer size needs to be increased.

[jasonwagner]

Great Idea Ski, I will try that thank you!!

[ski]

Cool!

 

Yeah, sometimes the little noise artifacts that can be produced by a too-small buffer size will be masked by the actual recorded audio content, but you might end up hearing them when monitoring in (say) a quieter environment, or when you isolate a particular track.

 

Ideally, you want to do this kind of test without any cables physically connected to the audio interface's input to eliminate the possibility of line noise or other static being picked up. If you were to do that, increase the amount of Change Gain by greater than 24 dB. Essentially, the idea is to boost the level of very low-level pops, clicks, or other artifacts to a point where you can see them easily (and, if you played back the silent tracks, hear them easily). Just be sure to lower the audio output a LOT just in case.

[jasonwagner]

We are good to go! Thanks guys, I feel much more at ease now. LOL

[David Nahmani]

Good to hear! Good luck!