Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Nov 11, 2018 19:29:20 GMT -6
Are all the advantages of higher frequency sampling rates for tracking and mixing lost when mastering down to a final product at 44.1Khz? Some are, but at least you're only causing the phase inaccuracies a single time. If the project originates at 44.1, then those errors are compounded multiple times -- through every filter or compressor.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Nov 11, 2018 19:37:00 GMT -6
I wonder how taxing it would be on the computer though. And I will need a disk space calculator to figure out the extra cost in disk space. In part, this depends on how well the DAW software and the plugins are written. But as a rule of thumb, you can figure that every doubling of sample rate will require a doubling (maybe a bit less) of CPU usage. You can look at your computer's activity monitors to see current load and get an approximate idea of whether you could go there. Hard disk space requirements will double as well, but storage is so cheap these days you'll hardly notice. Perhaps the biggest possible snag would be the number of tracks and what sort of disks you're working with. This may cause your system to cough even when you have plenty of CPU. 7200 RPM drives are pretty much a minimum and more and more people find that SSDs are a much better choice (as long as they're at least USB3 or T-Bolt 2).
|
|
|
Post by Bob Olhsson on Nov 15, 2018 13:43:22 GMT -6
A LOT of plug-ins sound better at higher rates. I bought a 6TB HGST helium-filled internal spinner and haven't even thought about disk space in a couple years. I use it for both audio and disk image backups of my two SSD system disks.
|
|
|
Post by Mister Chase on Nov 15, 2018 13:47:32 GMT -6
A LOT of plug-ins sound better at higher rates. I bought a 6TB HGST helium-filled internal spinner and haven't even thought about disk space in a couple years. I use it for both audio and disk image backups of my two SSD system disks. I feel the same. My questions are this: Do the plugs sound better because they are operating at higher rates, or because the source material is at a higher rate? I suppose you could sample up your material and go 441 vs 192 in the daw to find out.
Then, the plugins that oversample - do they still sound better with source material at higher rates?
It's all a bit difficult to truly A/B I would imagine. I am curious, though.
|
|
|
Post by Blackdawg on Nov 15, 2018 13:50:49 GMT -6
A LOT of plug-ins sound better at higher rates. I bought a 6TB HGST helium-filled internal spinner and haven't even thought about disk space in a couple years. I use it for both audio and disk image backups of my two SSD system disks. I feel the same. My questions are this: Do the plugs sound better because they are operating at higher rates, or because the source material is at a higher rate? I suppose you could sample up your material and go 441 vs 192 in the daw to find out.
Then, the plugins that oversample - do they still sound better with source material at higher rates?
It's all a bit difficult to truly A/B I would imagine. I am curious, though.
I know some plugins actually upsample material themselves. At least that used to be a thing..im struggling to remember what plugin i knew did this..very cpu intensive though. I'd save a little of column A and a little of column B though for this.
|
|
|
Post by Bob Olhsson on Nov 15, 2018 14:15:36 GMT -6
Many are worth upsampling with Saracon or RX. I hear the difference after they become 44.1x16 or even MP3s.
|
|
|
Post by Mister Chase on Nov 15, 2018 16:02:35 GMT -6
Many are worth upsampling with Saracon or RX. I hear the difference after they become 44.1x16 or even MP3s. I wonder if I should upsample with RX anything I mix that comes in at less than 88.2 That'll be interesting.
|
|
|
Post by Bob Olhsson on Nov 15, 2018 17:24:05 GMT -6
I use 96 because I've found plug ins that sound better at 48 and 96 than 88.2.
|
|
|
Post by Guitar on Nov 16, 2018 7:47:27 GMT -6
I've been listening to some John Coltrane records at 24/192 and they sound... ...really good. It's that sense of realism that I think someone mentioned.
With this kind of music I think the high rates make a lot of sense.
|
|
|
Post by drumhead57 on Nov 16, 2018 11:35:18 GMT -6
I like my Coltrane at 5.6 mHz DSD
|
|
|
Post by Guitar on Nov 16, 2018 11:41:34 GMT -6
I like my Coltrane at 5.6 mHz DSD good idea... hmm....
|
|
|
Post by Mister Chase on Nov 16, 2018 12:07:00 GMT -6
I just got the new Bill Evans Black Forest lost recordings album in 192/24
I would like to try against 441 to see what I hear. I wonder how many of those hd tracks are just up sampled. I'll have to throw SPAN on it.
|
|
|
Post by popmann on Nov 16, 2018 13:01:42 GMT -6
This is a sore subject for me....first, there are not any “upsampled” on HDTracks...they police that TOO well, IMO....as somehing recorded at 44 or 48 CAN be masterd analog and presented in HD and have tangibly better sound. They wont allow THAT....so something like John Mayer’s continuum is only presented at 24/44 on HDTracks, where internationally you can buy the acrual 24/96 mastering-the one used to cut the vinyl records and will be used to stream 96 as future streaming moves there.
Part of the reason I so recommend people use HD is theres not the one singular advantage. Its not JUST a truer to analog input capture. While “upsampling” alone doesnt really do much for any low rez tracks, it CAN be part of a process that makes better sound. No one points out that one of those “bogeyman” scenarios was that in the early 00s, many SACDs were actually from 48khz DAT and Protools sessions....which I would say delivering the 24/48 IS better than 16/44, particuly given that inexpensive SACD DACs sounded DAMN good, IMO, where that same inexpensive DAC playing CD was foul...and they missed that the purpose of SACD was so that Sony could “equally” archive their analog tapes and 50 different digital formats of IP. Literally why they invented DSD—not to make a consumer format. So, if Celine Dion or whomever recorded at 48khz—it deliver d that....96? Analog straight to DSD? All of the above “equally well”. So, there no “foul” to cry for them using a 48khz DAT as source for and SACD. Particularly givn that it was almost CERTAINLY mastered analog and capturd with the DSD, since theres no DSD digital mastering tools....
Fwiw....you can stream Tidal for free for a month. If you use the desktop app, it will unpack the MQA to 96khz without a special DAC (other than obviously one that can play 96khz)....there sound to me like theres more to MQA....it seems more spacious and open up top than the “allegedly same” 96 and 192 files sitting on my drive. Regardless, it sounds fab. Im actually paying them, depite saying I would never support streaming....but, streaming never had content I couldnt buy before....and the Eagles Long Road sounds great in HD....while its a mix of tape and 96 and 48 and 44....it was all mixed analog and mastered to 192 for vinyl originally....Frey being audiophile and they sold the vinyl direct rather than the Walmart deal....anyway—theyll likely eventually release it outside the $240 complete thing on HDT....but, for now, Im rediscovring it on Tidal.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Nov 16, 2018 13:56:17 GMT -6
I just got the new Bill Evans Black Forest lost recordings album in 192/24 I would like to try against 441 to see what I hear. I wonder how many of those hd tracks are just up sampled. I'll have to throw SPAN on it. That's the whole problem really. If your brain/ear/converter/monitor/room interface can't tell the difference, is it even worth checking it with an analyser? Genuine question.
|
|
|
Post by Mister Chase on Nov 16, 2018 14:43:07 GMT -6
I just got the new Bill Evans Black Forest lost recordings album in 192/24 I would like to try against 441 to see what I hear. I wonder how many of those hd tracks are just up sampled. I'll have to throw SPAN on it. That's the whole problem really. If your brain/ear/converter/monitor/room interface can't tell the difference, is it even worth checking it with an analyser? Genuine question.
It's very interesting. I wish I could somehow record an acoustic guitar simultaneously at 441 and 192 for instance. Same preamp signal split off. I suppose that is doable if I had two sets of identical converters. What I have to do currently is play something twice. Not good enough for a true A/B. If I simply down-sample with RX, that isn't the same either as many have stated the gains of higher rates make it down to a degree when down sampling versus just recording at 441(which I can't say I totally understand, because I would think a filter in a nice set of AD like the Lynx Aurora (n) could do the job, but maybe non real time digital filters do it better?)
I did however try a VSTi test. I used Abbey Road Studio Drummer and mixed it down from VSTi land at 441, 882, 96 and 192. What I found was that 192 was subtly wider and deeper. I mean, I had to listen for it. The real difference was the high hats/cymbals. Less harsh and spitty the higher you went. It was easily noticeable. I noticed from 441 to 882 a little, not much from 88 to 96, but definitely to 192. However, I am curious if this is due to other factors than simply better resolution.
My question is, at 192khz, are there more samples at 10khz(for example) per second than 441khz? Is it a more accurate high frequency content? It's interesting because I obviously don't need much above even 18khz on a recording, and I see a lot of noise the closer I get to 96khz on my recordings which I don't want. However the trade off may be more samples per frequency?
Some have argued it's the problems that arise at a rate like 192khz that cause smearing etc that we tend to find more pleasing.
This is probably the most mysterious of the audio facets to me. I am sure the Illuminati are involved.
For me, 192 sounded easiest on the ears and widest/deepest for VSTi. I am fairly certain that plugins will sound better here as well. However, I definitely would need to low pass every mix. Just sending a blank signal through the converter loop at 192khz shows the loudest part of the noise floor at 96khz. -60 dbfs.
|
|
|
Post by Mister Chase on Nov 16, 2018 14:47:48 GMT -6
This is a sore subject for me....first, there are not any “upsampled” on HDTracks...they police that TOO well, IMO....as somehing recorded at 44 or 48 CAN be masterd analog and presented in HD and have tangibly better sound. They wont allow THAT....so something like John Mayer’s continuum is only presented at 24/44 on HDTracks, where internationally you can buy the acrual 24/96 mastering-the one used to cut the vinyl records and will be used to stream 96 as future streaming moves there. Part of the reason I so recommend people use HD is theres not the one singular advantage. Its not JUST a truer to analog input capture. While “upsampling” alone doesnt really do much for any low rez tracks, it CAN be part of a process that makes better sound. No one points out that one of those “bogeyman” scenarios was that in the early 00s, many SACDs were actually from 48khz DAT and Protools sessions....which I would say delivering the 24/48 IS better than 16/44, particuly given that inexpensive SACD DACs sounded DAMN good, IMO, where that same inexpensive DAC playing CD was foul...and they missed that the purpose of SACD was so that Sony could “equally” archive their analog tapes and 50 different digital formats of IP. Literally why they invented DSD—not to make a consumer format. So, if Celine Dion or whomever recorded at 48khz—it deliver d that....96? Analog straight to DSD? All of the above “equally well”. So, there no “foul” to cry for them using a 48khz DAT as source for and SACD. Particularly givn that it was almost CERTAINLY mastered analog and capturd with the DSD, since theres no DSD digital mastering tools.... Fwiw....you can stream Tidal for free for a month. If you use the desktop app, it will unpack the MQA to 96khz without a special DAC (other than obviously one that can play 96khz)....there sound to me like theres more to MQA....it seems more spacious and open up top than the “allegedly same” 96 and 192 files sitting on my drive. Regardless, it sounds fab. Im actually paying them, depite saying I would never support streaming....but, streaming never had content I couldnt buy before....and the Eagles Long Road sounds great in HD....while its a mix of tape and 96 and 48 and 44....it was all mixed analog and mastered to 192 for vinyl originally....Frey being audiophile and they sold the vinyl direct rather than the Walmart deal....anyway—theyll likely eventually release it outside the $240 complete thing on HDT....but, for now, Im rediscovring it on Tidal. Now that is very intesting about HD tracks. It's a wonder they didn't just record at 96 anyway on something like Continuum which is a great record.
This whole sample rate thing has me interested in trying to find a conclusion. I have yet to do that.
I mean, in theory, 441khz shows info above 99% of human ears at 22.05khz. So if that were the only aspect, it would sound the same as higher rates. However, I seem to have found some difference in the rates. How can that be? The only thing I can imagine, as a layman, is that more samples per second in the high frequencies means a more accurate and better sound. Or, maybe its pleasing smearing. I have no idea. You have to think the filters on nice pro converters are good enough not to cause issues themselves. So it has to be something else, right?
The other thing that makes me believe this theory is that when I take the 192khz file and RX it down to 441, it almost sounds the same. Better than the regular 441 file. So its retaining the higher res up to 22.05k due to more sampling. The filter is fine at 441 but you are capturing less points per waveform.
It could also mean the filter in the ADDA isn't good, but somehow I doubt that.
I want to believe the basic premise that 441khz capture can theoretically sound the same as higher rates, but so far I have detected a difference.
All this without saying, 441khz won't ruin a record that is destined to be great. But I want to stack the deck in my favor.
|
|
|
Post by Guitar on Nov 16, 2018 15:38:22 GMT -6
I think at the higher rates the filters in your DAC have better time based resolution. The lower rates can have more pre and post ringing.
Apparently it is related to a mathematical principle called Gibbs Phenomenon.
That is my very simplistic interpretation. This is something I bang my head against from time to time. I love reading about it if anyone has any better or more complete ideas to share.
|
|
|
Post by Mister Chase on Nov 16, 2018 23:38:43 GMT -6
Hmmmmm. I know they could sing high, but I don't really know what the BeeGees have to do with it.
Yea I am very interested in this as well. It will really come back to what I like working at personally but it seems shrouded in debate.
However, if the ADDA caused ringing with the filters, why would down sampling from 192 retain a better sound? Are the digital filters just better? Surely they cause ringing as well...
|
|
|
Post by Bob Olhsson on Nov 20, 2018 19:03:47 GMT -6
Most recordings involve digital signal processing which typically involves multiple filters. High sample rates allow the use of more gentle filters. Unfortunately that doesn't mean developers automagically understand this but at least there is the potential for better sound. It's about how fragile the audio is when subjected to additional processing such as digital volume controls.
|
|
|
Post by b1 on Nov 21, 2018 13:03:38 GMT -6
Still happily doing 24/48k here. I did 96k for a while if it was heavy on plugins that sounded better there. Now-a-days, for me, plugins are just fillers.. Doing analog with a pinch of digital. I still do 192k if doing sound design since the higher resolution gives "easier/more" editing options. I just accept 48k and avoid the loop of trying to out-do somebody else's set-up.. Leaves time (space, money) to focus on what sounds good to me musically. Personally, I've come to the conclusion that it's more about the material than the process - as long as the sound is inspirational to me and genders other ideas from the root idea. I still dig some of my old 16/44.1 stuff. It was an inspiration that brought it into being and the nuances that came with it, well, it's painful for me to think of changing it and losing the whole spiritual feeling that comes back when I hear it; verbatim. Too much on my plate to even consider it anyway. Another life would be handy though .
|
|
|
Post by Bob Olhsson on Nov 22, 2018 11:35:23 GMT -6
This issue often gets into an argument of theory vs. practice. Most of us aren't writing our own code or building our own converter filters so we need to figure out what sounds best using the gear and software we already have.
|
|
|
Post by b1 on Nov 22, 2018 12:05:57 GMT -6
This issue often gets into an argument of theory vs. practice. Most of us aren't writing our own code or building our own converter filters so we need to figure out what sounds best using the gear and software we already have. I agree whole-heartedly. My thoughts were towards the listener. Usually if we do something that inspires us, there's a chance that it'll mean something to someone else - whatever the sample rate.. as long as it sounds good with fidelity of course. Only a minute portion of people would quibble about a process.. Quibble for a lack of a better term, but it's a legitimate argument on one level.. In the end, it's music with the creator or reproducer striving to perfect it for the benefit of others for a lasting effect.. Be happy and make music... which goes to your point.
|
|
|
Post by Mister Chase on Nov 22, 2018 12:19:16 GMT -6
Very true.
Thankfully I don't have huge 50 track sessions and with the pc power now, I can run 96 if I want to. I think that's pretty nice to be able to do on an end user type pc in this day and age.
|
|
|
Post by jazznoise on Nov 23, 2018 5:31:17 GMT -6
48/24. Happy without it sounds and couldn't really afford. To be losing adat inputs to 88/96 rates.
Re:the phase distortion of upper frequencies. Try putting some all pass filters above 10Khz. The mechanics of how the inner ear relates information to the brain implies that you shouldn't hear a difference,but I cant say I've tried it.
|
|
|
Post by Ward on Nov 23, 2018 7:43:53 GMT -6
48/24. Happy without it sounds and couldn't really afford. To be losing adat inputs to 88/96 rates. Re:the phase distortion of upper frequencies. Try putting some all pass filters above 10Khz. The mechanics of how the inner ear relates information to the brain implies that you shouldn't hear a difference,but I cant say I've tried it. So..., LOW PASS FILTERS SET TO 10K? That might end up sounding dull, lifeless or muddy. Even if most folks over 30 can't hear above 12K, you're still cutting out a whole lot of information. I routinely slope things down from 16K but only because that's all people really get with an MP3 or back in the day a cassette tape. And that's the sound most folks prefer anyhow. Bonus: Limiting bandwidth from 35.5hz to 16.1Khz gives you more headroom and the possibility for loudness. A heck of a lot of real estate gets chewed up from 20-35.5hz and from 16-20Khz
|
|