Although the discussion was recently mainly on the bandwidth aspect, that IMO should not imply we all think it is reason/cure for all that matters.
You as one of the big Hi-res supporters, please try to suggest a track of an album.
This has absolutely and nothing to do with cost, just and only with the question wether Hi-Res produces a better sound because of more HF, less envelope distortion, longer word length or whatever.
Hans
I'm afraid I won't even be able to give a meaningful listening impression due to the limitations of my tech.
I must be crazy - where did I express an assumption on anything? 384ksps was a reference to where ordinary HW capacity has come today. By posting that statement, there where no attached confirmation of that this is needed, better, worse or anything really. It was just that - info on current market situation for consumer A/D boxes.
(reading back now I see indeed that you made an own interpretation in #1452 of my basic factoid that you quoted)
//
Jakob2, you interpreted that I made an "implicitely given assumption (i.e. that sampling at Fs at 192 kHz or even above) solves any problem that might exist."
This is an erroneous interpretation. I never ment, nor wrote, any such thing. Maybe you need to be a bit mor cautious with your interpretation and use of other statements?
//
I agree with storage and bandwidth still decreasing in price, Hires is less problematic from a cost perspective. As long as we are talking subtle differences between protocols, another factor to consider is the increased difficulty of accurate LSB with higher bitrate 24-bit AD and DAs. I am not aware of any good test reports showing this with available 192 and 348 KHz sampling ADs.
The current crop of 48 and 96 KHz 24-bit ADs have excellent and stable LSB tracking...
Cheers,
Howie
At true resolution beyond 21 bit, I think that what LSB happens to end up with is of no importance any-longer - let it be random so the Dither aficionados are happy 🙂
//
//
ADC tech
Howie, what do you mean by current crop? Only reason I ask is I was confused until recently corrected on how SD ADC work and they all run the converter flat out and decimate to the desired rate. Do many people even make traditional audio ADCs any more?
It might be not that significant difference for a consumer with his 2 channel USB or whatever DAC, but for people recording music, other than the purist 2 tracks or limited multitrack, every step upwards from 24/48 is coupled with big investments. Changing to the latest DAW is only one part of the deal - if one works in the digital domain all the outboard has to be re-purchased too.
The fact that Joe Doe and his brother can now play 384k on their Win laptop doesn't mean that the whole industry will suddenly jump to 384k.
Here's one interesting indicative example - the digital EQ used in almost all mastering studios around the world and probaly on 80-90% of releases. Can you spot the 192kHz in the specs? 😉
EQ1 | Products | Weiss Engineering Ltd.
If 192k would have been THAT inportant, I bet that the particular EQ would have been upgraded to 192k long time ago.
Yes, I know, there are some out there who record in DSD, or 384k, but that kind of releases are only small part of the picture.
Working at Amazon, I responded to a "next big thing" employee contest with "Amazon Tracks" - an imaginary music streaming service where 8 tracks are provided for the user to mix down any way they'd like.
Of course, the "original" mix, eq, compression settings would be provided in a meta file so if the consumer did nothing, it would sound as the recording engineer originally intended - with all such processing happening in the player app.
One guy commented that he'd like to be able to self remix / re-effect his music purchases - the rest of the community was silent. (I'm sure most had no idea of what I was talking about).
Funny how this 384k provides a little more than 8 channels of 44.1k...for the same stream / download cost. I imagine it might take someone like Amazon to be able to talk the music industry out of the keys to the castle, so to speak. I also imagine forums where people talk about their specific mixes, eqs and compression ratios and how that effects "how it sounds". I suppose people could become "famous" for having developed a certain "mix" of some pop tune that many others found desirable.
Given we have infinite bandwidth, infinite post processing capability - I see this as only a matter of time - when someone else figures out this potential has been on the table, remains unused - and does something about it. Given the above, an 8 channel start would ultimately expand to unlimited. But the 8 channels at 44.1 - mix down any way you want - versus the same old poop but this time at ridiculous bandwidth; I wonder which you personally could make sound better?
Given the depth of interest in controlling all possible variables in sound reproduction I see here on DIYA, I'd think such an idea would be a hit. Apparently, Amazon didnt think so, as I never did get that email from some VP telling me "Go for it - you own it!".
Of course, the "original" mix, eq, compression settings would be provided in a meta file so if the consumer did nothing, it would sound as the recording engineer originally intended - with all such processing happening in the player app.
One guy commented that he'd like to be able to self remix / re-effect his music purchases - the rest of the community was silent. (I'm sure most had no idea of what I was talking about).
Funny how this 384k provides a little more than 8 channels of 44.1k...for the same stream / download cost. I imagine it might take someone like Amazon to be able to talk the music industry out of the keys to the castle, so to speak. I also imagine forums where people talk about their specific mixes, eqs and compression ratios and how that effects "how it sounds". I suppose people could become "famous" for having developed a certain "mix" of some pop tune that many others found desirable.
Given we have infinite bandwidth, infinite post processing capability - I see this as only a matter of time - when someone else figures out this potential has been on the table, remains unused - and does something about it. Given the above, an 8 channel start would ultimately expand to unlimited. But the 8 channels at 44.1 - mix down any way you want - versus the same old poop but this time at ridiculous bandwidth; I wonder which you personally could make sound better?
Given the depth of interest in controlling all possible variables in sound reproduction I see here on DIYA, I'd think such an idea would be a hit. Apparently, Amazon didnt think so, as I never did get that email from some VP telling me "Go for it - you own it!".
Last edited:
And if the tweaks here are OK with that as well, then as you say there is no problem🙂 In my experience I would state that for a delivery format 21 bits of resolution is sufficient for almost all recordings. Not for initial recording, but for consumer delivery.
Right! My comment reflects mainly the fact that I personally have not seen any dithered LSB waveforms for 192 or 348 KHz 24-bit ADCs...although they may be out there. Maybe they do settle out fast enough to provide good 21-24-bit accuracy, though it is difficult enough at 96k AFAIK.
Maybe I am just woefully uninformed...
Cheers!
Howie
If a "real" recording, i.e. with a mic, I think 21 is sufficient fo any stage, no?
But I suppose you think about added noise through a multi-ch mix?
I think live recording of live event with few mic - hifi ;-D
//
I think not...and thank you for visiting the retirement home...wait...I ain't retired! Dang you TnT!
Cheers!
Howie
Is there any advantage to higher sample rate during mixing and processing?21 bits in mix means you can maximize SNR in capture and maintain it in the mix, even if attenuated by a lot, should you desire. Spot SNR and ENOB (I'm being sloppy when units, apologies) on any one mic/track may be in the 16-18 bit range, with a mix needing 21+ to preserve then ensemble.
Edit-- extra bits are nice to not run into truncation errors, ScottJ.
*I'm not pretending to know whether there's any benefit herein or you end up having signal being dithered by local environmental noise, even such quiet places as Ed's.
Edit-- extra bits are nice to not run into truncation errors, ScottJ.
*I'm not pretending to know whether there's any benefit herein or you end up having signal being dithered by local environmental noise, even such quiet places as Ed's.
Yup, the two main advantages of running 24-bit tracking is lower summed noise in the final mix, but just as importantly ease of level setting. You always ask for a sound check from a musician, but once they get into it they can get carried away and peak levels can skyrocket. If the take was a real winner, you would hate to listen to playback and hear clipping! 24 bits lets you run levels down in the tracking and retain good headroom.
This principle applies to live sound as well, some sound engineers who do not monitor SPL become progressively deaf as a show progresses and levels can rise quite a bit. Especially among those engineers who balance a mix on the fly by potting an instrument up, as opposed to knocking the rest of the mix back and bringing up master gains which takes more time but preserves headroom and overall SPL in the room.
Cheers!
Howie
Okok 🙂 I ment 21 efficient/used bits. Surely 24b is the platform choice. I treat myself to set the knobs of my mic amplifier with an extra 4-6 dB attenuation so that afterwords, I can either use it as is or amplify the whole thing 6 dB by doing a 1b shift left. That preserves all bits but the LSB 🙂
Those 4-6 dB also takes care of the "heated spirit" moments 🙂
//
Those 4-6 dB also takes care of the "heated spirit" moments 🙂
//