Some DAW questions - panning, higher sampling/bit rate

[UstadKhanAli]

I know how to track, mix, etc. But my knowledge of DAWs is still lacking. Here are some questions:

 

1. Pro Tools recently switched to 32-bit float. I can now run sessions at a 32-bit float instead of 24 bits, and this sounds better to me. Recently, I noticed that I also have the option of increasing the sample rate as well during mixdown. My question is: If I track everything at 44.1kHz, is there any upside to mixing down at a higher sampling rate, like 88.2kHz or 192kHz, as I know can with Pro Tools?

 

2. Can someone please explain panning laws to me? I just realized that I can change the panning from the default of -3dB to less or more than that. Is there any reason I should do that? How would it impact summing to mono? I suspect that when I do this, it affects the pan when it is at 0 (straight up). Can someone explain this in layman's terms to me? When should one ever use, say, -6dB, for example?

 

Pro Tools 10.0.1, Mac Pro1,1, I forgot my OS in case this matters.

[WRGKMC]

The upside to upsampleing when mixing down is if you use mastering plugins.

You have to think in digital terms and how a wave is sampled. The way I remember it is this.

If you see a chart of a sine waves responce in an X/Y chart, in digital it would consist of dots,(separate measurements of voltage height and frequency in time)

in analog you'd see a solid line consisting of height in voltage vertically and and frequency in time moving left to right.

 

The waveform is sampled digitally. You know that. The sine wave is sampled in measurements both

vertically and horisontally. The height (dynamics) sample rate is your 16, 24, 32 bits. The width of the waveform

is your 44, 48, 96, etc. These separate measurements in the thousands per second are converted back to an analog

waveform by connecting and smoothing the dots. Like a CRT screen, the light beam moves so fast, the eyes dont see the blinking.

same with the ears with separate dots creating the sound.

 

If you take a digital photo, and enlarge it enough, you see the picture is made up of dots. When editing a photo and cropping it the boxes shades

change from one color to another through a series of half tone boxes along the otherwise sharp edges in normal view.

 

By enlarging the photo you can crop the edges much more accurately by removing single pixials, or even half pixials. When you reduce back down the

edit is sharp and accurate.

 

Same happens when you enlarge an audio file horrizontally. The cropping used by mastering effects will be more accurate and leave the uncropped areas

undisturbed so whan you reduce it back down, less data is destroyed, more original quality is preserved, and less noise/distortion from missing data is introduced.

 

It wont add any quality, but less quality is sacrificed using the plugins.

Question comes down to, Will up sampling make that big a difference in what you hear?

Theoretically, more data is preserved and you could analize it down and prove it mathamatically.

Will the primative ear drims be abole to hear such small amounts of data loss at such high frequencies hear it is the question.

My thing is Data loss is collective. The more times you process a signal even just up sampleing, and putting a waveform

under a microscope for surgery by a plugin has losses. If you want to retain quality you have to try it and hear if the loss is worth it or not.

If you cant hear a difference, its a waste of time, if you can you have to decide if its worth the difference. Do a file both ways with the same

plugin settings and A/B them in random order and see if you can hear a difference on a good playback system.

 

You may find recordin at the higher sample rate is more benificial. You just have to decide if the music recorded is worth the extra effort.

if the music was played by star musicians and its going to be a hit single, I'd record it at maximum sample rates so i'd capture all that can be captured.

I could always crop what isnt needed. If its my own music, I'd rarely need anything above 24/48 which is how I record.

[Zooey]

 

Pro Tools recently switched to 32-bit float. I can now run sessions at a 32-bit float instead of 24 bits, and this sounds better to me. Recently, I noticed that I also have the option of increasing the sample rate as well during mixdown. My question is: If I track everything at 44.1kHz, is there any upside to mixing down at a higher sampling rate, like 88.2kHz or 192kHz, as I know can with Pro Tools?

 

 

Native versions of PT have always used floating point, so the headroom advantages of floating point for real time processing and summing have always been there. The difference is that you can now track to 32 bit files and make sessions out of them. This does offer a few advantages for certain types of processing (like rendering).

 

As for "increasing the sample rate during mixdown," if you already tracked at 44.1, I have never personally heard any improvement from upsampling the final stereo mix. The reason I think that option is there is to give you flexibility if you need to deliver a master that is at a higher sample rate. You didn't need to wait until PT10 to try this, by the way. Upsampling was previously available using "export selected as files."

[CME]

IMO 32-bit float is a waste of space while tracking. Pretty much all converters these days are 24-bit fixed. So I'd leave that alone as 32-bit float will store a bunch of extra 00000's and use hard drive space. However once you're done tracking, you can switch to 32-bit float and gain more headroom through the mixer. That can be beneficial, but if you've gain staged properly shouldn't make much difference really.

 

I'll say they've covered the sample rate part.

 

Pan laws are as you think. As you move a sound from one side to the other, if you had no dip in level (0db pan law) it would get louder at center. The pan law can now be selected to be -2.5 (what pro tools used to use), -3, -4.5 and -6. I believe -6 gives you the theoretical best mono sum. But in most cases you'll hear a decrease in volume as a sound panned through center. -3 gives you the most natural sound when panned through center. And 4.5 is a compromise between -3 and -6. However IMO once you've started mixing you're gonna compensate with fader moves anyways. Still best to occasionally flip to mono and make sure everything still works no matter which pan law you choose.

[UstadKhanAli]

Thank you very very much for these answers!!!!!