|
Post by hio on Apr 12, 2019 18:34:08 GMT -6
When I took the aforementioned test I was on my Tablet but I had good headphones.
I carefully A/B them multiple times before committing and probably why I got most of them right; two though I picked the 320 kbps.
|
|
|
Post by NoTomorrow on Apr 12, 2019 18:37:59 GMT -6
Actually I have a new 10 core Imac Pro so Logic and Pro Tools see 20 cores/threads. It's ridiculously powerful. And I've been on Mac since 1997 so I won't be changing over.... although I am a little jealous of the performance some of you Windows guys can get. The problem for me is Acustica plugins, which are fantastic.... but at 96k are monsters. It only takes maxing out a single core to stop the transport with an error even though I'll still have 19 free cores just sitting there waiting for instructions. It's more a problem when chaining plugins on a single channel. And when chaining on the master bus obviously. PT cpu distribution in general is suspect. If you put 6 plugs on a single channel that 'almost' max it out and then create an Aux and feed it with that same channel and move half of the plugins to the Aux, your CPU usage goes way down. I wish they would make the routing cpu distribution 'smarter'. Are the Acustica plugs worth it 96k??, yes.... to me they are. So quite a bit of freezing and rerouting has to go on for me at 96k and I don't do every project at that sample rate, but I do hear a difference. Believe me, I didn't want to. This is my exact experience too with AA and my Mac Pro cheesegrater. I have been tracking/mixing at 96Khz lately and if you have good analog gear on the way in, it does sound more natural to me...subtly, but it’s different than 44.1 and 48. Still on the cheesegrater huh? What, you didn't like the garbage can lol There's supposed to be a new Mac Pro toward the end of the year. Maybe it will look like the Millineum Falcon or the Eiffel Tower and cost $7000 Incidentally Spindrift, I was the guy on Reverb about a week ago inquiring about your Kircher mic - thanks for responding to me. Then I went to SF for a couple of days and when I got home I saw a Stam SA-67 on Reverb and grabbed it. If I end up not digging it and yours it still available - I'll buy it from you
|
|
|
Post by spindrift on Apr 12, 2019 20:58:22 GMT -6
Oh, thanks! It just sold though.
I maxed out my cheesegrater and it is humming along strong...if it costs 7K, I’m gonna rock the cheesegrater for a while longer. It has loads of power...but yes, Too many acoustics plugs do hang it up. I like you aux channel trick. I’ll try that.
|
|
|
Post by NoTomorrow on Apr 12, 2019 22:38:17 GMT -6
Ah you sold it... ok, congrats. I know it was on Reverb for a little while there. So I was building some drum kits tonight in Pro Tools and doing some more testing with Acustica plugs. The difference is huge if you route channels with multiple plugins through 2 Aux channels (for some reason using 3 Aux channels is worse than 2). See this graphic: (tested at 24/44.1)
|
|
|
Post by jdc on Apr 13, 2019 5:41:45 GMT -6
I usually find that I notice differences over time, so quick A-B tests can often be misleading. I take my time with a thing, and eventually I have an epiphany, I notice some quality I like or I don't and then react accordingly. I haven't been able to use 96k often enough to know for sure if I hear any differences. I can identify with this. I need to really learn the sound of a thing inside and out over a long period of time, and once I've found what I don't like about it that's all I can hear. This goes for microphones. outboard gear, speakers, converters etc. The more experienced I become the more time I take to assess something and let my ears listen into it as deeply as I can. I'm still learning this skill and always will be.
|
|
|
Post by Martin John Butler on Apr 13, 2019 7:42:32 GMT -6
Exactly jdc! That's the thing with mics, amps, consoles, preamps, cables, all of it.
|
|
|
Post by Ward on Apr 13, 2019 20:29:43 GMT -6
I think we're all looking for 'more open and natural sounding'. Which is why we all like a GOOD Neumann U67 . . . .and to that end, the Audio Scape V-comp makes it even better, and well . . . if that's what 96 can add more to, well, it might be time to up the horsepower and RAM and SSDs and do the big sessions in 96K
|
|
|
Post by roundbadge on Apr 14, 2019 4:13:03 GMT -6
on the Burl mothership 96k sounds more effortless to me. that said I run 48 mainly because I need the extra dsp on bigger sessions. sounds fine either way
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Apr 14, 2019 11:32:55 GMT -6
Sorry to be late on this thread. I was at NAB earlier in the week and am just catching up. The first question I would have is "what's wrong with clinical"? When I record, I want to hear as close as possible just what hit the mics. Higher sample rates make this happen. If you take some time and a piece of graph paper, you can prove to yourself that higher frequencies (anything up to nyquist) are more and more inaccurate in terms of phase as frequency goes up. This would be true with sample rates up to infinity. But the higher the sample rate, the more accurate the phase. When those high frequencies are partials of some instrument, then the phase accuracy has (I believe) a big effect on the sense of localization of a signal. This is perhaps not as crucial in a pan-potted environment. But as soon as you use a stereo mic of any sort, it's going to make a difference.
Of course when you deliver at 44.1 or 48, you're going to have to downsample. But if the rest of the workflow is at those lower sample rates, the phase issues may be compounded in every plugin, every filter and ever gain control. I'm overstating a little, but I think this introduces a kind of mushiness. Some folks may like this effect and that could be why they like the lower rates.
96K and above place a real demand on a studio. Storage needs increase, CPU load multiplies. Issues in the monitoring system (crossovers, phase linearities and so on) are exposed. But if you can swing it, the results are noticeable.
|
|
|
Post by Johnkenn on Apr 14, 2019 12:42:44 GMT -6
Sorry to be late on this thread. I was at NAB earlier in the week and am just catching up. The first question I would have is "what's wrong with clinical"? When I record, I want to hear as close as possible just what hit the mics. Higher sample rates make this happen. If you take some time and a piece of graph paper, you can prove to yourself that higher frequencies (anything up to nyquist) are more and more inaccurate in terms of phase as frequency goes up. This would be true with sample rates up to infinity. But the higher the sample rate, the more accurate the phase. When those high frequencies are partials of some instrument, then the phase accuracy has (I believe) a big effect on the sense of localization of a signal. This is perhaps not as crucial in a pan-potted environment. But as soon as you use a stereo mic of any sort, it's going to make a difference. Of course when you deliver at 44.1 or 48, you're going to have to downsample. But if the rest of the workflow is at those lower sample rates, the phase issues may be compounded in every plugin, every filter and ever gain control. I'm overstating a little, but I think this introduces a kind of mushiness. Some folks may like this effect and that could be why they like the lower rates. 96K and above place a real demand on a studio. Storage needs increase, CPU load multiplies. Issues in the monitoring system (crossovers, phase linearities and so on) are exposed. But if you can swing it, the results are noticeable. I kind of meant “clinical” in the pejorative sense. Maybe sometimes strident. But I’ve also recorded in 96 and it sounded great. Tape is all the rage and all sorts of plugin emulations of now “inferior” gear to get all the non linear artifacts. I’m not convinced one way or the other necessarily. I still think we’d be hard pressed to find better recordings than the White album and Kind of Blue.
|
|
ericn
Temp
Balance Engineer
Posts: 14,917
|
Post by ericn on Apr 14, 2019 13:19:39 GMT -6
Sorry to be late on this thread. I was at NAB earlier in the week and am just catching up. The first question I would have is "what's wrong with clinical"? When I record, I want to hear as close as possible just what hit the mics. Higher sample rates make this happen. If you take some time and a piece of graph paper, you can prove to yourself that higher frequencies (anything up to nyquist) are more and more inaccurate in terms of phase as frequency goes up. This would be true with sample rates up to infinity. But the higher the sample rate, the more accurate the phase. When those high frequencies are partials of some instrument, then the phase accuracy has (I believe) a big effect on the sense of localization of a signal. This is perhaps not as crucial in a pan-potted environment. But as soon as you use a stereo mic of any sort, it's going to make a difference. Of course when you deliver at 44.1 or 48, you're going to have to downsample. But if the rest of the workflow is at those lower sample rates, the phase issues may be compounded in every plugin, every filter and ever gain control. I'm overstating a little, but I think this introduces a kind of mushiness. Some folks may like this effect and that could be why they like the lower rates. 96K and above place a real demand on a studio. Storage needs increase, CPU load multiplies. Issues in the monitoring system (crossovers, phase linearities and so on) are exposed. But if you can swing it, the results are noticeable. I kind of meant “clinical” in the pejorative sense. Maybe sometimes strident. But I’ve also recorded in 96 and it sounded great. Tape is all the rage and all sorts of plugin emulations of now “inferior” gear to get all the non linear artifacts. I’m not convinced one way or the other necessarily. I still think we’d be hard pressed to find better recordings than the White album and Kind of Blue. I think at this point it’s not about what is best overall, it’s what imperfect format works best on the project. The only rule is there are now set rules.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Apr 16, 2019 7:45:14 GMT -6
I kind of meant “clinical” in the pejorative sense. Maybe sometimes strident. But I’ve also recorded in 96 and it sounded great. Tape is all the rage and all sorts of plugin emulations of now “inferior” gear to get all the non linear artifacts. That description was certainly accurate when applied to that sorry first generation of 'DDD' CDs. We were still largely in the world of resister-ladder converters. They had all sorts of problems with foldover and jinky brick wall filters. There was a real bite in the upper mid and lower HF range. Dither was voodoo and not well implemented. Thank goodness we got past all that, and now 44.1 can sound very nice. My suspicion is that higher sample rates may expose other issues in the chain--interfaces, crossovers, tweeters and such. If they were designed around a 20K upper limit, it may be that the additional high frequency stuff is just a little too much. The resultant artifacts may be similar enough to that first generation digital that the term clinical (in your pejorative sense) could be a close fit. And I really do know what you mean about plugin emulations of inferior gear. I added an optional word-size reduction feature in all my second generation reverbs to let you add quantization error--just like older 'verbs. I think all that fizz sounds horrible and I'd never use it myself. But a fair number of people reported that it's just what they wanted. Go figure
|
|
|
Post by Johnkenn on Apr 16, 2019 16:32:15 GMT -6
Ok...ok...I might be coming around a little. Just starting to mix the first song and boy, seems like plugs respond better...and reverbs sound a lot better.
|
|
|
Post by hio on Apr 17, 2019 14:21:41 GMT -6
Could be the Wheaties you had for breakfast or the cocktail you were drinkin. Attachment Deleted
|
|