Ask Me Audio Ep. 15: Is There Any Quality Loss in Digital Copies?

In Episode 15 of "Ask Me Audio," Allen Sides explores whether or not there is any noticeable generational loss when copying from Digital.
Allen Sides is not only the owner of Ocean Way Recording, he is also one of the most respected engineer/producers in the music industry.
As an engineer/producer, Sides' has recorded over 500 albums and won 5 Grammy's. A brief list of some of the artists he has worked with include: Phil Collins, Green Day, Eric Clapton, Alanis Morrisette, Faith Hill, Trisha Yearwood, Wynonna Judd, Beck, Mary J Blige, Ry Cooder, Joni Mitchell, Frank Sinatra, Ray Charles, Count Basie, Duke Ellington. Ella Fitzgerald, John Williams, Jerry Goldsmith, Tom Newman, Andre' Previn, and Frank Zappa.
Got a question? Ask below in the comments and Allen might just answer your query in a future video.
For more information, follow Ocean Way Audio online at oceanwayaudio.com
On Instagram: / oceanwayaudio
On Facebook: / oceanwayaudio
Ocean Way Audio products are also available at Sweetwater: www.sweetwater.com/store/manu...

[@hansthijs]

Indeed. Tell it like it is, although you can’t explain what’s the reason. Well done. I like your experienced answers👍🏽

[@lmanna]

Very interesting take on this. I work with Digital Copies of Genome Sequences. The whole idea of these Genome Sequences is to compare them with other Sequences of the same organism to id SNIP .. i.e. 1 letter changes in the genome which will amount to 1 byte difference. These letter changes in the genome helps id changes in the genome of say cancer cells. If we had errors of the scale you describe in copies of digital audio the whole genome sequence comparison edifice would fall apart. However errors DO occur in copying data from a Hard Disk array to another or even worse from a medium (Hard Disk) to anohter (Solid State Drive or to CD-RW). To mitigate these errors the program that does the copies have error correncting protocols that would id errors in tne transfer and request another copy and correct the transfer. Even after the system confirms that the copy has completed without errors some engineers (the paranoid, like me ) would run a hash algorithm over the whole chuck of data on both the source and the destination and compare the results. If the Hash Numbers are the same we would be confident the copy is bit-by-bit exact copy of the original. I am not sure if these error-checking and hash-algorithms are available in ProTools and other Audio Software.

[@BrentLeVasseur]

The only thing I can think of is DACs make a huge difference. If you changed DACs between listening tests that could be the reason you are hearing a difference.

[@oceanwayaudio3029]

From Allen Sides: "Are there audible losses? My answer is Yes Yes and Yes but the degree of loss varies dramatically depending on the circumstance and equipment involved. Maybe the best way to explain this would be to define my experience mastering various albums and to point out a couple fairly blatant potential issues for loss. I should also say that my good friend Bernie Grundman made me aware of some of these issues. The first one is the difference between me sending my mixes online to the mastering house as compared to me bringing the actual mix hard drive to the mastering session.
Second would be the difference between using that original hard drive as compared to a memory stick. Here’s what I found. The original mix hard drive definitely sounds the best. The memory stick copy sounds noticeably worse and the mixes sent online also sound noticeable worse. I would say the losses are not huge but definitely noticeable. The qualities of the sound loss is smaller size, less wide and although the extreme highs are affected slightly there was also a loss of low-end presence. Some people might not immediately notice the loss unless they could directly compare the source drive to the other options.
I think a lot of people just assume it will be fine and never actually do a comparison. I have had situations where I have recorded large orchestras on various scoring stages which all had Neve 88-R consoles of which I also had in one in one of my rooms. We made copies from the stage’s internal hard drives and came back to my studio to mix. I had printed live 7.1 mixes during the session so I had a good reference point. We recorded at 192 32-bit resolution and l mixed down to the standard film format 48k 24-bit but I could directly A/B the high-res new mix output to what I had printed on the stage. The mixes I printed on the stage had more air and space. Clearly the 48k mixes would not sound quite as good but the high-res ones should have sounded very close; they didn’t and that was only one transfer to a hard drive and then brought over to my studio, same exact console.
The best way to hear these losses clearly is to first have a very accurate and defined monitor system, second is that the differences are certainly more noticeable if the quality of the recording is excellent and third that you are making this comparison at exactly matched levels. I have had all sorts of other loss issues over the years through various technologies but I will save that for another time!
Lastly I love Philips’ motto (who invented the CD) “Perfect Sound Forever.” Remember Philips wanted to use 32k sampling for the CD because the European broadcast standard was nothing above 15k so 32k was fine with them because that meant 15k bandwidth. Sony who was their partner wanted 50k sampling so they settled on 44.1. Obviously at the higher the sampling rate there are more samples at the higher frequencies, therefor more resolution. It’s not just bandwidth. Sony realized this, Philips didn’t. As far as forever goes I think etched stone tablets are it!"

[@BrentLeVasseur]

@@oceanwayaudio3029 Thanks! And that’s also why a lot of audiophile record labels today are recording in DSD 256 and then mixing in 32bit 768k, because DSD is a more natural digital representation of the analogue longitudinal waveform that is sound. Also as an audiophile myself, it’s a shame that your orchestral recordings at 32bit weren’t re-released as 32bit PCM or DSD 256/512 High res recordings. Since everything is a digital download today anyway and most have given up the physical media formats like CDs and Vinyl, why not re-release those recordings in high res formats for true audiophiles? That would be amazing. Cheers and thanks!