You are right.
I also noticed that I read JCX post wrong. You wrote very loud volume, which is probably correct, but once you notice the details at loud volume, you will also begin to listen for them at low volume, and they are there if your domestic noise level is low enough (hate my fridge!).
Nevertheless we can probably all agree that if recording studios where not using compression, they would be better at their jobs because they would actually do something. And there may be a possibility that the following increased level of detail would justify the increase in resolution.
Last edited:
Hi.
You win.
I will have a very hard time telling a properly dithered 16 bit 44k1 file from a 24 bit 96k one. There are differences, but it is very hard to tell them apart in a blind test. The only way you can tell them apart is if you are listening to the same track and constantly changing between 16 bit dither and 24 bit throughout the tune.
That being said, the dithered one is better at low volume levels in the track, while the non-dithered 16 bit is better at higher volume levels in the track. The 24 bit version seems to catch the best of both the 16 bit dither low volume detail level, and the regular 16 bit high volume detail level.
It is very late here now, maybe I will test with higher sample rate tomorrow.
Again: It is nearly impossible to tell any of the three apart unless you keep switching constantly. But I'd go for the 24 bit if there was a choice, good at all levels. The result might be different if I actually had a stereo here right now, still waiting for that !#¤=%!"=(& paintjob......
24 bit will sell better because the number is bigger. Does not matter how good 16 bit with dither is.
I don't care much about the sample rate, tho 48Khz might be nice because it's a simple resample from 96KHz, the studio standard.
Just give me the recordings without all the dynamic compression. They can do it on SACD and DVD-A. Why not better masters for download?
I don't care much about the sample rate, tho 48Khz might be nice because it's a simple resample from 96KHz, the studio standard.
Just give me the recordings without all the dynamic compression. They can do it on SACD and DVD-A. Why not better masters for download?
The musicians want compression, compressed source gets better reviews (even in the so-called "high end" journals) and more radio play. We're not even big enough to be a niche. Too bad, uncompressed 16/44 is amazingly good- in ABX, I couldn't tell it from 24/96 when the levels were set properly.
Compression does sell records, it always has. But there is at least some awareness in the musical community of the loudness wars. Vise Bob Dylan's rant of a years back.
Let the masses have their 12dB dynamic range. Just offer us more, we'll pay for it. If it starts to sell, then more people will offer it. Audiophile pressings and special CDs have sold for years, to a small market. "Better" downloads would be super easy and a money maker, it seems.
Let the masses have their 12dB dynamic range. Just offer us more, we'll pay for it. If it starts to sell, then more people will offer it. Audiophile pressings and special CDs have sold for years, to a small market. "Better" downloads would be super easy and a money maker, it seems.
I'd like the compression to do the vanishing act, earlier I listened to Pendulums In-Silico album and it quite literally hurt my hearing for the day, 10+ hrs and my hearing is still not right, chuffing headache still aint gone either.
No wonder kids say 'wot' and look blank when they pull their ipood plugs from their ears.
Am easy on the sample rate part
No wonder kids say 'wot' and look blank when they pull their ipood plugs from their ears.
Am easy on the sample rate part
The petition has loads more signatures now (552!) thanks to this link! Its clear that lot of people are fed up with compressed music.
Petition for Better Sound | Stereophile.com
Petition for Better Sound | Stereophile.com
You would have to listen for harmonics on hard notes at the beginning of the sound on dithered 16 bit. At the end of the sound/transition to low level reverb on non dithered 16 bit. Nearly impossible to hear, but it is possible. In both cases there is some small amount of information missing.
Does anyone know any other wav comparison tools than FourierRocks? It only supports 16 bit.
Thank you for this, will investigate when I get home. A pity it does not support 32 bit, was thinking of trying that as well.
There was no conversion, I rendered. I thought it would be difficult/impossible to provoke any difference with existing music in my library, so I dusted off a remix I made some years ago, the interesting bit is that this was an experiment with what I think of as daisy chained distortion limiting with automation. Every sound will distort a little bit differently and therefore produce harmonics, the notes also vary in velocity and this will also make slightly different distortion, even at low volume passages.
Do not know if it makes sense, more along the lines of: Keep your friends (dynamics) close, but keep your enemies (distortion) closer.
Will post results no matter what I find out. Unsure about the depth of the investigation attempts though. I have 2 small kids and a wonderful wife that do not get enough attention.
I am sorry about the delay in posting results.
Tried the Diffmaker program you suggested SY. It starts up alright, settings seem fine, run the analysis between the 16 bit dither and regular render, the files themselves are near identical. When starting the analysis the program goes through both files ok, the program eats memory though, peaking at 1500MB, but when running comparison and merging it crashes, saying there's a problem with the files being to different. Tried varying settings and compatibility modes without any more luck.
The FourierRocks is also very interesting, fired it up, ran analysis for 3.5 hours on the 16 bit files before it crashed when I tried opening Opera to post here. I've set down quality from Good to Average, memory usage still varies from 600MB to around 2500MB. CPU usage is reasonable at around 25%. Will let it run overnight to see if it manages to finish.
... Anyone else want to give it a go? Can share the files with dropbox. This is just to provoke out the differences in dithered/regular 16 bit and hopefully also compare 16 bit to 24 bit. There should be no/very little difference, but there seems to be a lot of difference in how harmonics are handled.
Edit:
Will also try with regular converted files, 24bit to 16 bit regular and dither.
Tried the Diffmaker program you suggested SY. It starts up alright, settings seem fine, run the analysis between the 16 bit dither and regular render, the files themselves are near identical. When starting the analysis the program goes through both files ok, the program eats memory though, peaking at 1500MB, but when running comparison and merging it crashes, saying there's a problem with the files being to different. Tried varying settings and compatibility modes without any more luck.
The FourierRocks is also very interesting, fired it up, ran analysis for 3.5 hours on the 16 bit files before it crashed when I tried opening Opera to post here. I've set down quality from Good to Average, memory usage still varies from 600MB to around 2500MB. CPU usage is reasonable at around 25%. Will let it run overnight to see if it manages to finish.
... Anyone else want to give it a go? Can share the files with dropbox. This is just to provoke out the differences in dithered/regular 16 bit and hopefully also compare 16 bit to 24 bit. There should be no/very little difference, but there seems to be a lot of difference in how harmonics are handled.
Edit:
Will also try with regular converted files, 24bit to 16 bit regular and dither.
Last edited:
You don't resample with dither, you use it after downcoverting the bit depth. Like going from 24 bits to 16 bits. There is no need for dither when resampling from 96khz to 48khz as all I am doing is filtering out high inaudible frequencies.
While I agree with the statement in principle, it does not mean the benefits of going higher than 16/44.1khz aren't there.
If you say the situation is unsettled, then how can you recommend anything until it is? Secondly, if you have the storage space, then why downcovert the digital audio when you don't need to. Dither IMO is strictly a Redbook CD thing, as you are trying to fit a high amount of data into a restricted space. If you are not restricted to the Redbook standard, then there is really no need to downcovert, and then add dither. It is a unneeded process. Aside from that, dither is no panacea. When you add something to something that is not native to that something, it changes it. That change in my experience is it adds a "graininess" and a "hardness" to the sound. Do you hear this when you have no comparison to the orginal source? Probably not.
You are probably right - but what is the point? We are living in the days of cheap digital storage, and Bluray disc, both which negates the need for downcoversion and the use of dither. IME it would be far better to record in 24/96khz, and deliver it to the end user at that same bit and sample rate. Less(processing) is more has always been my policy. If I don't have to downcovert, I don't.
There are a lot of forums catering to musicians and the science of recording. Many of these are frequented by professional recording engineers, and these engineers often make source material for a song freely available so that home recording enthusiasts can mix the song themselves.
If you are serious about this endeavor, then you should have no problem raising $5K (maybe via kickstarter) and you could use that money to hire an engineer to mix a song exactly as you want. In other words, you could "produce" the song precisely as you see fit. My guess is you could find an engineer in a studio for $50 to $100/hour and he could mix a song to your specs in 5-7 hours, and you could get 5 songs of various types mixed to your specs for a few thousand dollars.
These 5 songs would then be your calling card to make a case. If you succeed, the benefit will be obvious and getting others on board will be easy. If the benefit isn't obvious, then you still learn something.
So, troll the various recording forums, find some potential source material you like, do a kickstarter and show the world how good this can be.
if you can't raise the money via kickstarter, then that's also tells you asking a record company to invest in this puny market (as determined via kickstarter) is silly. If you raise the money quickly via kickstarter, then that says there's a solid market for this stuff. If you raise the money quickly AND you succeed in producing the type of music you envision, then you have some credibility to approach some large acts and ask if you can do this on a grand scale.
If you are serious about this endeavor, then you should have no problem raising $5K (maybe via kickstarter) and you could use that money to hire an engineer to mix a song exactly as you want. In other words, you could "produce" the song precisely as you see fit. My guess is you could find an engineer in a studio for $50 to $100/hour and he could mix a song to your specs in 5-7 hours, and you could get 5 songs of various types mixed to your specs for a few thousand dollars.
These 5 songs would then be your calling card to make a case. If you succeed, the benefit will be obvious and getting others on board will be easy. If the benefit isn't obvious, then you still learn something.
So, troll the various recording forums, find some potential source material you like, do a kickstarter and show the world how good this can be.
if you can't raise the money via kickstarter, then that's also tells you asking a record company to invest in this puny market (as determined via kickstarter) is silly. If you raise the money quickly via kickstarter, then that says there's a solid market for this stuff. If you raise the money quickly AND you succeed in producing the type of music you envision, then you have some credibility to approach some large acts and ask if you can do this on a grand scale.
All - The petition letter has now been revised. Let me know what you think, ideally through the facebook group. Please try not to burn me at the stake if there are grammar errors! Thanks
http://www.change.org/petitions/stop-the-loudness-wars-and-release-high-definition-music-downloads
P.S – 858 signatures now – carry on spreading the word. If you haven’t already signed, please do so.
http://www.change.org/petitions/stop-the-loudness-wars-and-release-high-definition-music-downloads
P.S – 858 signatures now – carry on spreading the word. If you haven’t already signed, please do so.
If I were you, I would erase this line from the petition. Nobody in the music industry thinks 16 bit is high resolution, and no major record company is interested in vinyl - they are done with that and quite frankly all of the old technology. I would personally advance the use of the Bluray disc, downloads, and streaming in place of CD and vinyl since that is where the masses are headed in terms of high resolution audio.
With sales dropping like a rock with CD, the record companies are eager to abandon it as soon as possible(my sources tell me at the end of 2013). The market for vinyl is so small, the record companies are not going to invest a quarter(maybe a penny) towards it. If you look at sales trends(and the record companies do), downloads, streaming and Bluray disc are pretty hot in terms of sales of music(with Bluray being pretty far behind downloads and streaming, but capable of much higher resolution).
You cant please everyone!
I have actually been burned to the ground in the Sound on Sound forum for even asking for anything higher than 16 Bit. They are telling me the exact opposite and have told me that 24 Bit offers no improvement over 16 Bit, if its produced correctly in the first place. They have even said that even asking for 24 Bit would discredit the petition.
Obviously I don't agree with them, as I believe I can hear the differences.
As for as dropping vinyl, if they are going to press a vinyl copy anyway this is no more work for them, as the Masters are generally done in the digital world anyway. I've also left in the physical format, as its still a huge market, but yes I think that downloads are the future.
I have actually been burned to the ground in the Sound on Sound forum for even asking for anything higher than 16 Bit. They are telling me the exact opposite and have told me that 24 Bit offers no improvement over 16 Bit, if its produced correctly in the first place. They have even said that even asking for 24 Bit would discredit the petition.
Obviously I don't agree with them, as I believe I can hear the differences.
As for as dropping vinyl, if they are going to press a vinyl copy anyway this is no more work for them, as the Masters are generally done in the digital world anyway. I've also left in the physical format, as its still a huge market, but yes I think that downloads are the future.
done that blind with higher sample rate and good dither like I'm advocating?
what I am trying to warn against is the apparently popular thinking that 24/44 or 24/48 is a meaningful improvement for consumer music delivery format - "hi res" should be 1st about higher sample rate - that is the edge of known psychoacoustics that is "too tight"
with good dither even 16/44 doesn't have a noise problem with real music, played back in real consumer's living rooms or even good home theater rooms which seldom reach NC-20-25
but Nyquist at 22 kHz is a tight squeeze for known human audio perception frequency limits, practical considerations of anti-image/reconstruction filtering
by increasing sample rate, with dither, you also push down the perceptually weighted noise floor, "for free"
what I am trying to warn against is the apparently popular thinking that 24/44 or 24/48 is a meaningful improvement for consumer music delivery format - "hi res" should be 1st about higher sample rate - that is the edge of known psychoacoustics that is "too tight"
with good dither even 16/44 doesn't have a noise problem with real music, played back in real consumer's living rooms or even good home theater rooms which seldom reach NC-20-25
but Nyquist at 22 kHz is a tight squeeze for known human audio perception frequency limits, practical considerations of anti-image/reconstruction filtering
by increasing sample rate, with dither, you also push down the perceptually weighted noise floor, "for free"
yawn, still doesnt address volume control, which was the entire reason I mentioned it
and can you supply a single valid reason for performing an additional dither when you can just release the 24/96 file and not have the problem to begin with?
I dont understand... why would you want to remove real bits to put fake bits back?
and can you supply a single valid reason for performing an additional dither when you can just release the 24/96 file and not have the problem to begin with?
I dont understand... why would you want to remove real bits to put fake bits back?
Last edited:
I don't object if 24/96 is on offer - but 1st ask for the 96
companies are offering 24/44, 24/48 as "hi res" - ever hear of the Beatles? - any guess what the master tape dynamic range of that era was?
24/44, 24/48 should be denounced as consumer fraud
there is no issue with volume control as has been explained - multiply it out, send the top 24 bits to the existing 24 bit DAC in most every current digital audio player - like the software already does
companies are offering 24/44, 24/48 as "hi res" - ever hear of the Beatles? - any guess what the master tape dynamic range of that era was?
24/44, 24/48 should be denounced as consumer fraud
there is no issue with volume control as has been explained - multiply it out, send the top 24 bits to the existing 24 bit DAC in most every current digital audio player - like the software already does
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.