Topic: SACD versus CD test file available

[Petrus]

The test file 16-24_testi.wav was made to compare 24/96 (originally SACD) to 16/44.1 CD-quality is now available at http://hosted.filefront.com/Jullepoika/

The size of the file is 470 MB and format is 24/96 WAV.

Download also the 16-24_testi.txt which explains how the file was made.

The basic idea is to embed random 30 second 16/44.1 sections into a 24/96 file and find out if it can be heard.

You must have a system capable of clean 24/96 WAV reproduction to utilize this file. Typical download time is several hours, so making your own test file could be faster. The .txt file explains how to do it.

[johnlynn]

Just to be clear SACD is not 24/96 or even a PCM encoded format. SACD uses DSD encoding @ 1bit/5.6Mhz. This test is actually DVD-A versus CD.

[Frank in JC]

That sounds like a ridiculous test.  Using the same technique, it's likely I could insert snippets of 320kbps MP3 encoded material into an uncompressed 16/48 stream without an experienced listener noticing anything amiss.

[Petrus]

The original file is sold as 24/96 file, but it is originally from SACD recording (judging from the noise spectrum). So I rather call this SACD versus CD test. Or an expose how relativelly inferior material is sold as 24/96. Defenetly not DVD-A versus CD. If it would make much difference.

Basically the setting is the same as in the recently published SACD-CD test, where people could not hear the difference. Just that with this file you just listen, no need for switching systems and levels matching.

And Frank in JC: you are welcome to make a 16/48 versus 320kbps MP3 test. If people can not say which is MP3 and which is 16/48, then MP3 is good enough and can not be told apart from 16/48. Would that be so awfull? Considering the popularity of MP3 that would almost be good news!

[johnlynn]

The original file is sold as 24/96 file, but it is originally from SACD recording (judging from the noise spectrum). So I rather call this SACD versus CD test.

Hold on that SACD recording originally came from live musicians. So lets be fair and call this Live Musicians versus CD.

If I record a vinyl record to PCM 16/44.1 Can we do a vinyl versus cd test tomorrow.

[Petrus]

Live musicans versus CD would require the precence of live musicans... That kind of tests were done in the eighties with symphony ochestra playing in one hall and live monitoring done in another. Signal was either straight wire analog or via 16/44.1 ADA conversion. No difference. On this file we have some sections in SACD (converted to 24/96 by the recording company), some in CD resolution. If you can not tell them apart, then there is clearly no audible difference...

If you want to compare vinyl and CD, you can take the analog signal from vinyl live and put it thru 16/44.1 ADA conversion. Need a comparator for this. This was done already in the eighties, the test person Igor Tiefenbrun (of Linn fame) could not tell the difference.

link: http://www.bostonaudiosociety.org/bas_speaker/abx_testing2.htm

[BC]

I think John is right, technically this is a DSD/SACD>24/96 PCM vs CD test. For a true comparison I think one should use un-converted native high-res material for comparison with 16/44.1 CD. I think that material sourced at 24/96 should ideally be used, not DSD, especially b/c DSD has large amounts of high-frequency noise that does not exist in 24/96 PCM.

But I guess this is the best this person could do with the material they had at hand, one just needs to be aware of the details.

[Petrus]

Two things about this test file:

- Inspiration comes from the AES SACD-CD test, so this basically duplicates that without the need for extra equipment. http://www.bostonaudiosociety.org/explanation.htm

- The original 24/96 file I used is sold as a "Studio Master" file by Linn. So this also compares something that is marketed and sold as super-hi resolution file vastly superor to CD (their marketing idea, not mine). I would rather have used a true 24/96 file, but I did not have any available. If somebody has one with truly great sound and dynamic range, please make a similar file from it, or send the original 24/96 to me and I'll make one.

[Frank in JC]

If people can not say which is MP3 and which is 16/48, then MP3 is good enough and can not be told apart from 16/48. Would that be so awfull?

If the MP3 snippets could not be readily detected by an experienced listener, it doesn't mean they cannot be told apart.  It only means the differences could not be detected in the context of that particular experiment.  It seems ill-conceived to me in the first place.  I don't think I would notice a fleeting difference between 16-bit uncompressed PCM and 320kbps MP3 when listening to music.  Through long-term listening however, the differences become more apparent.