|
Post by sozocaps on May 15, 2017 22:10:38 GMT -6
My best friend has recorded 20 albums ... real albums and they ALWAYS record at 48k and the last record went to Andy Sneap to mix and he was very happy with the tracking. What I hear is more energy and air at 48k VERY LITTLE but it's there and at his recommendation thats where I started recording and have NO regrets. Lastly who listens to CD's ? it's all other formats now which can natively transfer from your sample rate and 48k works good for music videos etc....
|
|
|
Post by aamicrophones on May 15, 2017 22:51:41 GMT -6
I agree with Jcoutou. I'm planning on trying 96 kHz next time I begin any new sessions. I've heard plug-ins can sound better at that sample rate. I'm unfamiliar with higher sample rates when bouncing. When I bounced my first 96 kHz session, it automatically converted to 48, is that normal? At, 48khz the filters are causing the high end frequencies to "fall over a cliff" at 70db/octave starting at 24khz. At, 44.1 khz the filters are causing the high frequencies to "fall over a cliff" at 70db/octave starting at 22.05 kHz. At, 96khz the filters are causing the high frequencies to "fall over a cliff" at 70db/octave starting at 48khz. To my old ears 48khz/24 bit sampling in my RADAR converters sounds as good as coming back from tape but I have less tape hiss and no low frequency bump from the record head. However, 96khz seems to sound more like the sound coming off the mixing console before it went to tape. The low frequency bump in an AMPEX 24 track was different than the bump from a 24 Track Studer or MCI 24 track. This was due to the different gap and inductance of the different maker's playback heads. That' why we always laid 1khz, 10khz and a 100hz tone on master 2" tapes that were moving from one studio to another; this way the low frequency EQ could be compensated to normalize the LF bump between different tape machines as well as the HF EQ could also be normalized. I did some guitar overdubs here at 96khz for a project that was being mixed in Sweden but the RADAR system drops from 24 tracks to 12 track at 96khz. Cheers, Dave
|
|
|
Post by EmRR on May 15, 2017 23:38:25 GMT -6
I do! My HHB CD player/recorder at home kicks the crap out of every other converter source I have in that location, apples to apples, 44.1/16 files from any converter. I'll also connect a MBP output to it via lightpipe to get it's converter benefits.
|
|
|
Post by javamad on May 16, 2017 6:09:25 GMT -6
I use 48k for all my sessions due to it being the standard for video. You never know when something will be requested for sync to video. We live in an audiovisual world :-)
I also plan to look at 96k but haven't got round to it yet. When I think my room, front end gear and monitoring are up to it, I will probably make the move.
|
|
|
Post by sozocaps on May 16, 2017 8:18:33 GMT -6
I agree with Jcoutou. I'm planning on trying 96 kHz next time I begin any new sessions. I've heard plug-ins can sound better at that sample rate. I'm unfamiliar with higher sample rates when bouncing. When I bounced my first 96 kHz session, it automatically converted to 48, is that normal? At, 48khz the filters are causing the high end frequencies to "fall over a cliff" at 70db/octave starting at 24khz. At, 44.1 khz the filters are causing the high frequencies to "fall over a cliff" at 70db/octave starting at 22.05 kHz. At, 96khz the filters are causing the high frequencies to "fall over a cliff" at 70db/octave starting at 48khz. To my old ears 48khz/24 bit sampling in my RADAR converters sounds as good as coming back from tape but I have less tape hiss and no low frequency bump from the record head. However, 96khz seems to sound more like the sound coming off the mixing console before it went to tape. The low frequency bump in an AMPEX 24 track was different than the bump from a 24 Track Studer or MCI 24 track. This was due to the different gap and inductance of the different maker's playback heads. That' why we always laid 1khz, 10khz and a 100hz tone on master 2" tapes that were moving from one studio to another; this way the low frequency EQ could be compensated to normalize the LF bump between different tape machines as well as the HF EQ could also be normalized. I did some guitar overdubs here at 96khz for a project that was being mixed in Sweden but the RADAR system drops from 24 tracks to 12 track at 96khz. Cheers, Dave GREAT info thank you... Maybe ill switch to 96k I have no issues with space or processing I do not do much mixing because I track correctly and play mostly acoustic and rock music. John
|
|
|
Post by Ward on May 16, 2017 9:48:41 GMT -6
I'm happy we've dumped the whole 88.2 idea, for the most part. Plug ins with Pro Tools seem to intensely dislike 88.2 Khz. Anyone have any ideas why?
|
|
|
Post by EmRR on May 16, 2017 10:10:28 GMT -6
I really only work at 88K2.....
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 16, 2017 13:43:10 GMT -6
I've heard plug-ins can sound better at that sample rate. There are some differences. A type of filter that's commonly used (biquadratic) will have slightly different characteristics at different sample rates. Most of the action around the passband is pretty consistent, but the remainder of any cut extends to nyquist. The higher the sample rate, the more gentle that part of the slope will be. There's more to it than that of course, and a large part depends on whether the filter is done in single or double-precision floating-point math. The difference between 44.1 and 48K is minor and quite hard to hear. It's probably not worth the hit of sample-rate conversion. But higher rates will reduce the effect of anti-aliasing filters in the capture phase and will give you a little bit less permanent schmutz in the recording.
|
|
|
Post by Bob Olhsson on May 16, 2017 15:02:10 GMT -6
I've heard surprising differences in older plugs between 44.1 and 48. Some sound worse at 88.2 than they do at 48 and 96! My guess is that choices were made at 48.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 16, 2017 15:15:03 GMT -6
I've heard surprising differences in older plugs between 44.1 and 48. Some sound worse at 88.2 than they do at 48 and 96! My guess is that choices were made at 48. Doesn't surprise me. Everything: delay lengths, filter coefficients, reverb time coefficients and more all must be computed for each sample rate. For a long time, developers worked on the mistaken assumption that 44.1 and 48K were the only sample rates that would ever come about. So everything was calculated based on just one rate and the assumption was made that the 10% difference wouldn't be noticed. You obviously noticed. This wasn't just true of older plugins. Even devices like the 480L and others of its generation were based on 44.1 and just stretched to 48. To the best of my knowledge, the 960L was the first Lexicon device that actually accounted for that, since it had to run from 44.1 to 96K. Nowadays, a plugin may have to go all the way from 44.1 (maybe even 32K) all the way up to 384K. There's no way to avoid the math then.
|
|
|
Post by Bob Olhsson on May 16, 2017 15:23:31 GMT -6
A lot of converters also used a single set of filters for both 48 and 44.1. There's what digital audio might have been and then there is what we actually were offered.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 16, 2017 16:56:51 GMT -6
A lot of converters also used a single set of filters for both 48 and 44.1. There's what digital audio might have been and then there is what we actually were offered. I do remember that first generation of CD players. Resistor-ladder converters, brick wall filters, no oversampling. Any quiet passages were just awful. It took a while to get there. Sigma-delta converters and noise-shaping took 44.1 to a higher level than I thought it would go. But more bits and higher sampling rates have driven noticeable problems considerably lower.
|
|
|
Post by svart on May 16, 2017 18:35:22 GMT -6
It's all about the converter.
Some will sound better at 44.1/48 than another at 88.2/96, etc..
That's The nature of hardware design..
It's really just about what you need balanced against what you're willing to put up with.
|
|
|
Post by Bob Olhsson on May 16, 2017 19:11:49 GMT -6
It's all about the filters.
|
|
|
Post by jeremygillespie on May 16, 2017 19:54:23 GMT -6
The digidesign 192's sound their best by far at 44.1
We did tests years ago and we were all pretty surprised by it.
|
|
|
Post by massivemastering on May 16, 2017 23:28:50 GMT -6
All I can submit is that if you can't make one of the most amazing, most natural sounding, audiophile-celebrated recordings of all time at 44.1kHz, it isn't the fault of the sample rate.
Same goes for 48k, 88.2 and 96.
|
|
|
Post by jazznoise on May 17, 2017 4:34:49 GMT -6
All I can submit is that if you can't make one of the most amazing, most natural sounding, audiophile-celebrated recordings of all time at 44.1kHz, it isn't the fault of the sample rate. Same goes for 48k, 88.2 and 96. I'd agree there. There are some things that can be worthwhile with using various sample rates, but the bottom line is 44.1K/24 on paper is extremely high fidelity that goes well beyond what the human hearing system is capable of. There may be technical errors, but they'll never represent the majority of issues in modern recordings.
|
|
|
Post by Bob Olhsson on May 17, 2017 5:59:58 GMT -6
I ran tests recording a live rhythm section and found the final sample rate converted mixes from 96k sounded better than the songs recorded a few minutes earlier at 44.1. Harman ran tests for multiple analog conversions in series and concluded there was a need in live sound systems for sample rates above 48k. If you are simply recording and then playing back audio, you probably won't hear much, if any difference but that isn't what we do.
|
|
|
Post by EmRR on May 17, 2017 8:09:19 GMT -6
All I can submit is that if you can't make one of the most amazing, most natural sounding, audiophile-celebrated recordings of all time at 44.1kHz, it isn't the fault of the sample rate. ...but as said previously, maybe the fault of the rest of the implementation. I submit no one will make a natural sounding recording of a drum kit and cymbals with a MOTU 2408mkIII at 44.1, but might at 88K2 or 96.
|
|