More > JRiver Media Center 24 for Windows
high quality music (hdtracks) is it worth it?
DJLegba:
--- Quote from: dtblair on March 26, 2019, 07:14:41 pm ---HDtracks is the worst. Here is proof they are upsampling
--- End quote ---
Good info. HDTracks knows there's a demand for all that old material, but since they're "HD", they don't sell 44/16. An alternate source is 7digital, where you can get 44/16 and sometimes higher resolutions if you prefer.
Larry - The three albums you mentioned were recorded in 1965, 1971, and 1977. The 1965 recording would have been the "cleanest", as it wouldn't have been multi-tracked and mixed in the studio with equipment that would look pretty simple today. I'm sure you listened to other albums, but in most genres other than jazz and classical even new recordings are usually heavily compressed and in many cases computers have replaced musicians. You're right though - production makes the biggest difference.
RD James:
--- Quote from: larryrup on March 26, 2019, 04:24:55 pm ---The recording/production/mastering is the number one certain ingredient in a wonderful sounding recording.
--- End quote ---
Absolutely.
--- Quote from: larryrup on March 26, 2019, 04:24:55 pm ---No degree of resolution can help a poor recording
--- End quote ---
Yes… but it's also important for people to understand that sample rate is not resolution.
--- Quote from: larryrup on March 26, 2019, 04:24:55 pm ---Listening for bit depth/resolution requires some training/practice. I now can generally tell any file at 256KB or lower, but at 16/44.1 (and I include lossless flac here) and 320KB it is not as easy...I think I can tell may 50% of the time through my home speakers and maybe 70% of the time through headphones but a wonderfully engineered recording blurs the difference. Listening in a car with just about any (including pricey optional) stereos I don't believe you would ever tell the difference between a CD and 320KB file. Further, higher resolution files in a car makes no sense to me. It's just too noisy
--- End quote ---
You are talking about bit-rate in lossy compression here, not bit-depth.
Bit-rate reduction via lossy compression - especially with a codec like MP3 - can be audible. Even high bit-rate MP3 has artifacts which are inherent to the codec - which is why it should not be used any more. AAC produced using a good encoder (I use Apple's via QAAC) is far superior.
I can pass a 320k MP3 vs FLAC blind test with 100% certainty. I haven't done an ABX test for these yet, but I use unrestricted VBR AAC encoding for portable devices, which tends to produce files about 1/3 to 1/2 the size of the equivalent FLAC, and I've yet to find a track where I've noticed the compression. I still keep the FLAC originals on my media server though, where storage is not a concern.
--- Quote from: larryrup on March 26, 2019, 04:24:55 pm ---Hi res files at 24/96 do not seems to distinguish themselves very much from 16/44.1. 24/192 do distinguish themselves more often, but NOT consistently, nor most of the time. My guess is they are noticeably better ~40% of the time....again for what I sampled which was certainly not chosen with any criteria as to what might have mattered (like age of recording, newly remastered...etc..... what I listened to/tested was what I came across and had interest in). When hi res was noticeably better to my ears, was it because of remastering? Source material? Resolution? I would think if it was the resolution, I'd hear a noticeable improvement more often than I did.
--- End quote ---
If it sounds different (better) it's almost certainly sourced from a different master. You need to compare 16-bit 44.1kHz tracks produced from the same "high-res" source.
--- Quote from: larryrup on March 26, 2019, 04:24:55 pm ---I once read (it was a while ago) that bit depth and not the resolution represents the biggest improvement in hi res audio. I didn't hear that or at least, can't draw that conclusion.
--- End quote ---
Well bit-depth is probably the closest thing to "resolution" in a PCM file. But if the track is dithered, literally the only difference bit-depth makes is how quiet the noise floor will be.
With 16-bit audio, the noise floor should be inaudible in most setups to begin with. If it does become audible, it's likely that your system is loud enough to cause hearing damage within minutes.
Note: this is different from the playback device's bit-depth.
If you are using digital volume control, particularly if your device is paired with an amplifier that has too much gain, it's possible that there will be a clear difference in noise between a 16-bit and 24-bit DAC.
But that is a completely separate issue from the bit-depth of the file being played.
--- Quote from: larryrup on March 26, 2019, 04:24:55 pm ---One question for the learned people on this topic about the higher file size of hi res files: Does ALL of that additional file size fall below and above the human hearing range? If not it's quite easy to conclude it adds to the depth of the recording. If so, I'm inclined to believe what improvements I hear are NOT attributable to the bit depth or resolution.
--- End quote ---
Yes.
One of the things a lot of people incorrectly assume is that sample rate can affect the "timing" of a signal, but that is not how digital audio works.
I recommend watching the whole video to get a basic understanding of how digital audio works, but here is a short demo showing how sample rate does not place restrictions on timing: https://www.youtube.com/watch?v=cIQ9IXSUzuM#t=1254
High bit-depths and sample rates do matter for music production.
If you are recording audio at 44.1kHz, which can capture audio frequencies up to 22.05kHz, and something in the room emits a higher frequency noise at say 30kHz like some of the examples posted above, that will cause aliasing if it is picked up by the microphone rather than being filtered out.
The aliasing from that 30kHz noise would then be folded back into the signal and cause distortion in the audible range (if I'm not mistaken, it would be at ~14.1kHz).
Recording at 96kHz can capture all frequencies up to 48kHz without aliasing.
That 30kHz noise would be present in the recording, but would not cause aliasing and distortion inside the audible range since it is a valid signal for a 96kHz recording.
You can then produce a 44.1kHz track from that 96kHz recording which has everything above ~21kHz filtered out digitally so that there is no aliasing.
It's a similar situation with bit-depth.
If you record at 16-bit, and something plays too loud, it may blow out the sound and cause it to clip. Clipping is a very audible and ugly distortion - a loud crackling noise.
If you record at 24-bit you have an additional 8-bits of headroom, or about 48dB. Now that very loud noise will not cause clipping distortion in the recording.
The actual dynamic range of a mastered music track is nothing close to the ~96dB of 16-bit audio, so there's no need to use 24-bit for a delivery/playback format, but it matters for recording because it helps prevent clipping distortion.
But people saw that studios were using higher bit-depths and sample rates than were being delivered to us as a playback format and basically thought they were holding back something from us.
And someone else came along and saw the potential for selling this as "high resolution" music they could charge more for.
BillT:
--- Quote from: RD James on March 27, 2019, 03:34:09 am ---High bit-depths and sample rates do matter for music production.
If you are recording audio at 44.1kHz, which can capture audio frequencies up to 22.05kHz, and something in the room emits a higher frequency noise at say 30kHz like some of the examples posted above, that will cause aliasing if it is picked up by the microphone rather than being filtered out.
The aliasing from that 30kHz noise would then be folded back into the signal and cause distortion in the audible range (if I'm not mistaken, it would be at ~14.1kHz).
--- End quote ---
While it's arguable that the greater dynamic range is useful in recording / post processing, you should not be getting aliasing due to out of band components in any competently designed digital system, whatever the sampling rate. The band limiting input filter and reconstruction filter are essential parts of the system and the input filter will remove any out of band audio signals - that's one of its main functions.
Of course, this isn't true of badly designed systems which deliberately omit the filter, but such systems have no place in a good recording setup.
RD James:
--- Quote from: BillT on March 27, 2019, 08:23:45 am ---While it's arguable that the greater dynamic range is useful in recording / post processing, you should not be getting aliasing due to out of band components in any competently designed digital system, whatever the sampling rate. The band limiting input filter and reconstruction filter are essential parts of the system and the input filter will remove any out of band audio signals - that's one of its main functions.
Of course, this isn't true of badly designed systems which deliberately omit the filter, but such systems have no place in a good recording setup.
--- End quote ---
Yes, it would be unnecessary with an ideal system, but it's really the main example where you might want to use a higher sample rate.
The same thing applies with certain plug-ins that are used to process the audio - some perform better at higher sample rates.
The point is that there is (sometimes) a good reason to use higher sample rates and bit-depths in production, but despite that, there's no need for the end product to be delivered at those same bit-depths and sample rates.
You aren't losing anything by receiving music at a lower bit-depth or sample rate than the "master track" - the track was only at those bit depths and sample rates to avoid potential issues which are specific only to its production, not playback.
larryrup:
Thanks much for all the info. Learning a lot. Thanks especially for the MP3 info and straight talk on why production may need head room where the consumer would not. It's a strange world where there is still a lot of strange talk about this. Some follow up questions if I may.
I somehow assumed there can only be one master. In my house I cite my wife as an example. There can be several? Clearly many hi res files sound different from lossless files of the same material, and if I understand this, it is due to a remastering from another master? I'm not sure I understand why there is more than one master and how they got here. It makes sense in an analog world to record to multiple tape machines (although I never thought about that), but in the digital world doesn't seem needed? You need backup but I would think that's it.
Do the marketers of these re-releases offer the CD (or lossless download) remastered or is only available in the hi res product? Starts to get complicated.
Took a pause and did a bit a internet searching on one of the albums I referenced by Jackson Brown. The HDTracks site offers only the 24/96 release from 2005. This is the only relevant description: "This hi-resolution release adds a new warmth and depth to the recording, bringing out the nuances of Browne’s intimate, honest performances." I guess they want me to assume this means remastered? jeasss. They sell this without liner notes for $19 and change. Further research on the Wiki page for the album states it was re-released many times. In 2005 it was remastered, in both 5.1 audio with bonus tracks and the original 41 minute release in 24/96. I guess the later is what HD Tracks is selling. Neither makes reference of the master used which most certainly would have been tape.
On to Amazon where they oddly sell the CD labeled "import" from 2008 for $9.48. Amazon makes no mention of mastering. One of the user comments reads "This is a 1st generation transfer and it sounds muddy and flat. I had ordered this at the same time I ordered The Pretender and Late for the Sky, both of which have been remastered"
It didn't take a rocket man to figure out why the record business went south quick and deep. Does not look like lessons were learned.
Appreciate all the info.
LarryRup
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version