44.1 or 48k is there much difference?

[laboroflove]

i can have more space on my computer and all and more plugins etc.do you all think sonically when doing just rock n roll type music that 44.1 is just fine.?i know the higher the sample rate the more analog it sounds but 88.1 is out of the question.lol.my comp cant hang with all the overdubs and stuff.im now recording at 48k at 24 bits.but its frustrating when im readdy to bounce down if i have lots of plugings on the bounce handler cant keep up.anyway ive down all the computer stuff to maximize performance and even the pro tools oil change they call it.do the plugins ins not work as well at 44.1 or what?anyway thx

[5454stevef]

I've never been able to tell much, if any, difference between recording at 48 vs 44.1 Khz, but then I have not done any critical comparisons between the same signal recorded both ways, and on top of that, my hearing is not what it used to be in the high end due to my advanced age. : ) In theory, at least, all you get from 48 khz vs. 44.1 is a couple Khz extension in high end response. Most people can't hear it anyway, nor do I have any mics that can record that high a frequency. On top of that, my monitors can't reproduce it.

Some knowledgeable people recommend using sample rates that are evenly divisible by 44.1 for material that's going to end up on CD, because current CD technology ultimately requires downsampling to 44.1 Khz before burning the file to CD. The idea being that the floating point math done during conversion is much simpler and less prone to errors. Again, I've not done the legwork to see if this makes an audible difference with current technology, but it seems that it could to someone with those golden ears I'm always reading about. I've always just assumed that the highest available sample rate is best, but that may not be true when downsampling to 44.1 Khz.

I'd be very interested in hearing pros vs. cons on this matter from anyone who has done the legwork.

SF

[Tarktones]

The basic answer is that you shouldn't choose 44.1 or 48 based on a difference in quality between the two but rather what the end format will be.

If it's going in a movie or on DVD, do it at 48 (or multiples thereof). On a CD, do it at 44.1 (or multiples thereof).

[dpd]

The sample rate conversions from 48 to 44.1 can be a bit messy (IIRC, they up-sample to a common multiple and then down-sample), but it's just a matter of computational power to do it well. But, I'm with Tarktones on this one, pick the sample rate based on the end format.

BTW, here's a link to a nice SRC. r8brain (http://www.voxengo.com/product/r8brain/)

Here's another link to SRC comparisons (http://src.infinitewave.ca/)

[laboroflove]

so maybe i could record the song at 48k 24 bits and when i run my mastering on the song upsample to 88.1.88.1 i coul;d prob handle just one or two tracks.then goto 44.1 16 bits.why do you say going from 48k to 44.1is kinda messy?i dont understand.
or would it be better to record at 44.1 then when i master goto 88.1 or hec just stay at 44.1 but bob katz told me the l2 loves high sample rates .

[DAS]

Most (almost all?) sample rate converters these days upsample to a very high sample rate before downsampling back to the desired rate. They work a lot like A/D converters, which also initially sample at very high rates and then downsample to the desired storage rate. So these days there is very little if any advantage to the whole evenly divisible thing.

DVD audio (in fact most audio for video) these days is at 48kHz. I would go ahead and record at that rate if that were my destination. That way you will not need to sample rate convert at all. Likewise if your destination is audio CD go ahead and sample at 44.1. I'd be really surprised if you heard a difference between the two.

As for the L2 liking high sample rates...Compression is one of those processes that does benefit from higher sample rates (again, the difference between 44.1 & 48 is not consequential). Most units will upsample to a higher rate even if the data coming in is at 44 or 48. I do not know the background of Bob's comment about it specifically, but if it indeed performs better (for whatever reasons) at higher rates then that is something you can consider if you want to use one. But it doesn't necessarily apply across the board.

[Tarktones]

In regards to upsampling for certain processes....

Remember, once something is captured in digital format, there is no way to restore information that was not captured initially recorded. Upsampling from 48k to 96k means you're now using twice as much space to represent the same exact data. The in-between sample points are approximations of the 48k wave. You have made no inherent improvement in fidelity.

Upsampling may help with the accuracy of some plugins, however there are plugins that upsample internally (take a look at some of the newer UAD plugins modeling the Neve gear) and then bring it back down to whatever the host is running at.

The idea is to record at sample rates higher than the finished product so that the audio is at maximum fidelity all the way until it has to be compressed for the consumer medium.

I personally believe you're better off doing as few sample rate conversions as possible. If you're going for a CD and your rig can't handle 88.2 or 96, just keep it at 44.1k/24bit and then just dither to 16 bit when mastering. Remember, upsampling will not restore what was not recorded.

[laboroflove]

thx every1 and bob katz said pretty much the same.he said ityll prob be better to stay at 44.1 cause he dont like the SRC on the pro tools le.he said if i was sending it off to get mastered i should record at 48k then let the mastering guys src converter take it to 44.1. but i dont send nothing to get mastered.i just do it myself and check on a few different speakers how it translates then print it and on to the next project.i find most clients remember the experience of the project more than gearhead gaga that they dont understand anyhow. but ill admit i am a gearhead lol