Tracking and mixing at 88.2/24bit...anyone doing it?
- Thread starter darthjujuu
- Start date
KeithTidd
Robot Penis.
track in 88.2 > convert the session to 44.1 (i should know the name but i think it's "save copy in" in pro tools)
I usually edit in 44.1, because i can't edit at 88.2 at home. Honestly i don't think it makes a huge difference what sample rate you're at when editing though, it only makesa difference what you track at ( and it could matter what you mix at, but it doesn't have to).
BLUElightCory
Member
The difference is there either way, especially in the high end (cymbals, etc.) - It WOULD probably be easier to notice a difference in a less-crowded arrangement though.
Ah, ok, I didn't know there was a quick way to convert the whole batch to 44.1, so I figured you would wait until you were done editing since I assumed you had to export them individually - thanks for the info
it doesn't seem like so much of a hassle to track the session at a high sample rate... then comp it, down-sample, clean up the project and delete unused files (the high resolution ones) and then proceed editing/mixing at 44.1. the point seems to be better converter performance, in some situations. this compromise seems totally logical as the extra disk space used gets freed up once you're done comping, and you've already downsampled before mixing so the heavy cpu load is irrelevant. hrmm o_0
No offense but thats what most people say when they don't have good enough gear to hear the difference, or take advantage of the difference...
Thats like saying you couldn't hear the differences between tracking a metal band straight to the converters vs. tracking to a well maintained A800 Studer. Thats just plain absurd. Harmonic content and resolution are real and the bandwidth that you have to capture them has a big impact on the final product. regardless of it being a 16 bit 44k format.
Same reason you can almost always spot the difference between an ITB mix and one thats done on a real console.
Most high end converters i.e. Lavry, Mytek, Radar... are designed around higher sample rate recording and have analog stages with the ability to capture harmonic content that are as high as 100k hz in the frequency spectrum. Normally if your converter sound is better at a lower sample rate its more likely due to it being of an inferior design or having less that stellar analog paths. If the analog stages were not designed to have a bandwidth of at least 50-60k hz, recording at a higher sample rate is only going to yield collecting more noise then the upper harmonic content of the signal.
my 2 cents
Fox Mulder
The Truth Is Out There
Sorry for the bump but I felt the need to share.
I recently started tracking at 88.2/24bit. The differences so far have been way beyond my expectations.
The entire frequency range of the source till the Nyquist frequency has to be processed by anti-aliasing filters (A/D) and reconstruction filters (D/A). Perfect filters aren't yet a reality. All commercially available converters, especially the low/mid-grade ones generate aliasing, oscillatory transients, and high-frequency phase shift.
The Nyquist frequency for 88.2khz is 44.1khz. This allocates the worst anomalies from imperfect filtering well beyond the audible band of human hearing. The only audio bandwidth that will ever matter is 20kHz and below. Digital audio sampled at 88.2kHz will have less aliasing and phase distortion in that critical frequency band than audio sampled at 44.1kHz will.
The cheaper your converters, the bigger the likely benefit of using a higher sample rate.
I recently started tracking at 88.2/24bit. The differences so far have been way beyond my expectations.
The entire frequency range of the source till the Nyquist frequency has to be processed by anti-aliasing filters (A/D) and reconstruction filters (D/A). Perfect filters aren't yet a reality. All commercially available converters, especially the low/mid-grade ones generate aliasing, oscillatory transients, and high-frequency phase shift.
The Nyquist frequency for 88.2khz is 44.1khz. This allocates the worst anomalies from imperfect filtering well beyond the audible band of human hearing. The only audio bandwidth that will ever matter is 20kHz and below. Digital audio sampled at 88.2kHz will have less aliasing and phase distortion in that critical frequency band than audio sampled at 44.1kHz will.
The cheaper your converters, the bigger the likely benefit of using a higher sample rate.
Machinated
Member
Apparently around 60kHz would be the optimum but as basically nothing is designed around that 48 or 88 would be your best bet.
TheWinterSnow
Den Mørke Natt
If you were going to keep the final sampling rate there for the final product, 88.2/24 would be totally worth it. I say this every time a sampling thread comes up but the thing is unless you are rocking some high end converters aliasing begins at half the Nyquist frequency, or a quarter the sampling frequency. Most audio above this become white noise. At 44.1K, the aliasing frequency is 11KHz. This is why it becomes very popular to high pass high gain guitars @12KHz. This is also why analog sims have to be used to darken up OHs and Room mics on drums to kill some of the white wash sound. Higher sampling frequencies are not plagued by this problem. At 88.2K, half the Niquist frequency becomes 22.05KHz, meaning, that no aliasing is taking place in the audible range only from 22.05KHz to 44.1KHz. The result is pretty interesting to say the least.
I did this test once on high gain guitars. Not only was the 88.2KHz sample rate brighter, the high frequencies where very musical and I did not need to high pass. The 44.1KHz sample frequency however, was darker and the upper mids and highs where fairly rancid and needed a LPF to take all that garbage out. In the end it leaves you with a lackluster mix. I noticed as well that "digital" sound was long gone at higher sampling rates.
Still none of this will matter when you downsample anyway as you will loose all of that extra resolution and will pick up the aliasing from the algorithms used to resample the audio. The only reason that you would really want to record anything higher than 44.1KHz is if you are going to keep the final format at that higher frequency for high quality listening. I would love to get the computer power so that I could record my band in 5.1 88.2/24. Blue Ray quality playback motherfuckers. It would be cool to sell high quality Blue Ray discs to the fan base. Mudvayne did their album "Lost and Found" in regular CD quality and DVD quality (48/24) double sided disc, it was pretty cool.
I did this test once on high gain guitars. Not only was the 88.2KHz sample rate brighter, the high frequencies where very musical and I did not need to high pass. The 44.1KHz sample frequency however, was darker and the upper mids and highs where fairly rancid and needed a LPF to take all that garbage out. In the end it leaves you with a lackluster mix. I noticed as well that "digital" sound was long gone at higher sampling rates.
Still none of this will matter when you downsample anyway as you will loose all of that extra resolution and will pick up the aliasing from the algorithms used to resample the audio. The only reason that you would really want to record anything higher than 44.1KHz is if you are going to keep the final format at that higher frequency for high quality listening. I would love to get the computer power so that I could record my band in 5.1 88.2/24. Blue Ray quality playback motherfuckers. It would be cool to sell high quality Blue Ray discs to the fan base. Mudvayne did their album "Lost and Found" in regular CD quality and DVD quality (48/24) double sided disc, it was pretty cool.
To me recording at 88.1 or 96khz makes perfect sense due to various improvements. Shorter latency at tracking, better or more true frequency response of the signal due to shifting most aliasing errors into the part of the audio spectrum where we don't perceive any information and an additional 2x oversampling during ITB mixing.
I haven't yet had a chance to try it out yet but some people over at Gearslutz said that they had great results converting sample rates from 88.2 to 44.1 using iZotope 64-bit SRC, retaining quite a lot improvements from using 88.2 throughout the whole production into their finished masters.
I haven't yet had a chance to try it out yet but some people over at Gearslutz said that they had great results converting sample rates from 88.2 to 44.1 using iZotope 64-bit SRC, retaining quite a lot improvements from using 88.2 throughout the whole production into their finished masters.
id like to hear a test from someone with better converters then me (motu clocked by presonus) program drums to a short clip , record DI guitars and bass at 96khz or whatever... reamp with a real amp (and good tone) , use more then one mic and make sure they are in phase, then use your favorite SRC on the DI tracks, change your samplerate in your DAW and reamp again at the lower samplerate. when i did it for myself i definitly heard a difference, amp distortion is a very complex "tone" and i think the different samplerates capture it differently (not better or worse in my case) but for me...the extra CPU headroom isnt worth it, ive been recording at 48 lately and planning on changing back to 44.1 once the studio slows down a little bit
