I'm too old to tell the difference, I have to leave that to you youngsters.
But I always rip my SACDs to 24x88.2 dsf anyway, just because I want the best sound possible for what I paid for (still play the discs direct also.)
QQ Survey Time!
- Thread starter Mr. Afternoon
- Start date
Help Support QuadraphonicQuad:
But I always rip my SACDs to 24x88.2 dsf anyway, just because I want the best sound possible for what I paid for (still play the discs direct also.)
jimfisheye
2K Club - QQ Super Nova
- Joined
- Jan 8, 2010
- Messages
- 3,555
Yeah then buy be some unit with at least 8 channels of DSD DAC! Needs to be at least Apogee quality that I'm used to. (Probably the only option is a Prism unit.) The transcode nulls down to almost 100db with the original. (There are cases of the same master being put to both DSD and PCM that you can compare like this.)
You're correct though! I do make some exceptions to that disclaimer I made that I should have mentioned! The other one is I will downsample any 192k program I get to 96k. This is another one I tested. 96k program upsampled/downsampled 100 times nulls perfectly with the original.
I guess the point I was trying to make (valid or not) was I avoid the crude conversions done for convenience. Like using an AVR with mystery on the fly conversions between PCM and DSD. (Most of those things only have PCM ADCs in them, folks.) Or doing a DSD to PCM transcode straight to 24 bit. Or to a different sample rate (and thus not an actual transcode). (eg. Foobar conversions with the default setting of 16 bit, 44.1k output.) Letting ripping apps rough split program at video frame chapter breaks resulting in clicks. All the gross stuff that is clearly audible. So apologies if you're not doing any of those things! That's what I was thinking about when I made the comment but I should have been more thorough.
So yes, if I was being completely thorough I'd keep copies of the DSD files and the 192k PCM files that come along. I'd still keep my Apogees and listen to the DSD content transcoded to PCM though. And I also believe 96k PCM audio holds a complete copy of any audio.
If you look at the converter designs between DSD and PCM, it's more like comparing serial data transmission with parallel (respectively).
I rip SACD (DSD) files as DSF and keep them as DSF whenever possible. I don't know if I can hear the difference between the straight DSF and the FLAC conversion; but I don't feel like spending the time converting anything if I don't have to. The only time I convert DSF to FLAC is when I want to fool with the audio stream in Audacity.
A few of SACDs that come to mind where I converted to FLAC:
- Company soundtrack...it needed a lot of special EQ to my ears
- American Woman...several tracks supposedly have swapped channels
- Days Of Future Past...I messed with the volume of the rears on various tracks
Ah. I've gotten help to edit DSD as DSD whenever I've needed it, so that makes sense.
True.
jimfisheye
2K Club - QQ Super Nova
- Joined
- Jan 8, 2010
- Messages
- 3,555
A broken analog clock is right twice a day.
With that thought in mind, one might accuse me of looking at, let's call it, a "ringer" example of a DSD to PCM transcode that matches a PCM copy. And either nulls down to absolute zero or at least nulls down to -100db.
That's logical.
I'm going to say that there is no ringer example where you can accidentally get two audio sources to even null down to -100db unless they are truly the same down to almost every last bit short of the noise floor.
Heck, just take two copies of the same file and lower the volume of one of them by 0.1db. That null test result has audio at -20db and is clearly audible music through laptop speakers! Most people would fail an A/B listening test to identify one of the files sounding different with up to a 0.5db volume difference between them! But you clearly hear the difference and see it on the meter with only 0.1db change between identical files.
When you can get a null result that is a little meter wiggle at -100db, that's the same audio source! Yes, something happened with a couple bits in the noise floor range. But it's the same audio and you'll never hear a difference in a million years. Not even with Adams or custom B&W speakers in the most perfectly constructed listening room.
And that leads me to...
When you ARE hearing obvious gross differences, it's something else messing with you! And again, a broken null test isn't coincidentally correct twice a listening session. Two audio sources with audible differences will never come close to nulling.
You can try some of this stuff yourself if you tend to believe no one and only trust your own experiments. Take a file, duplicate it, lower the volume of one of them by only 0.1db, flip the polarity and null them. What do you see and hear? Take two copies of something again with identical unity levels. Sample rate convert one of them. Now do it again back to the original sample rate. So 2x the lossy conversion there. Now null them. A little wiggle around -90db, right? Silence unless you turn your speakers up to very dangerous levels.
That's what I'm falling back on with this. If you actually hear something degraded, it's something else going on. DSD to PCM transcode isn't this destructive lossy piece of **** kind of thing!
With that thought in mind, one might accuse me of looking at, let's call it, a "ringer" example of a DSD to PCM transcode that matches a PCM copy. And either nulls down to absolute zero or at least nulls down to -100db.
That's logical.
I'm going to say that there is no ringer example where you can accidentally get two audio sources to even null down to -100db unless they are truly the same down to almost every last bit short of the noise floor.
Heck, just take two copies of the same file and lower the volume of one of them by 0.1db. That null test result has audio at -20db and is clearly audible music through laptop speakers! Most people would fail an A/B listening test to identify one of the files sounding different with up to a 0.5db volume difference between them! But you clearly hear the difference and see it on the meter with only 0.1db change between identical files.
When you can get a null result that is a little meter wiggle at -100db, that's the same audio source! Yes, something happened with a couple bits in the noise floor range. But it's the same audio and you'll never hear a difference in a million years. Not even with Adams or custom B&W speakers in the most perfectly constructed listening room.
And that leads me to...
When you ARE hearing obvious gross differences, it's something else messing with you! And again, a broken null test isn't coincidentally correct twice a listening session. Two audio sources with audible differences will never come close to nulling.
You can try some of this stuff yourself if you tend to believe no one and only trust your own experiments. Take a file, duplicate it, lower the volume of one of them by only 0.1db, flip the polarity and null them. What do you see and hear? Take two copies of something again with identical unity levels. Sample rate convert one of them. Now do it again back to the original sample rate. So 2x the lossy conversion there. Now null them. A little wiggle around -90db, right? Silence unless you turn your speakers up to very dangerous levels.
That's what I'm falling back on with this. If you actually hear something degraded, it's something else going on. DSD to PCM transcode isn't this destructive lossy piece of **** kind of thing!
Yeah, it's really only in theory. I do like to keep everything in DSD tho, just in case.
jimfisheye
2K Club - QQ Super Nova
- Joined
- Jan 8, 2010
- Messages
- 3,555
That's fair! If money were no object I'd probably buy some Prism DSD converters. (Apogee doesn't make any. Korg?! Haha, no. Does RME have a DSD interface?) The thing where Sony created an alternate digital language for the same end result to compete in a kind of dirty way colors the whole thing too. So I transcode and listen to PCM copies. And it's going to sound better that way than native DSD would on a cheap system. (Or converted on the fly in some AVR.) There's loss and then there's just silly loss!
But I kind of really blew that first comment, didn't I!
Then I'd need a patchbay or switching for home theater use. (Between the PCM unit and the DSD unit.) Analog switch/patch connections, oh my! Want to start a patchbay argument?
What's it supposed to be? 0.1v drop per connection?
JediJoker
Audio Engineer/Enthusiast
Ripping an SACD requires a compatible device (Blu-ray player/PS3), network connection, computer, USB thumb drive, and software (which at one point required either command line knowledge or Java be installed on the computer). That's a lot of complexity.
By contrast, with a Blu-ray Disc, everything is handled on a computer. All you need is an optical drive that will read Blu-ray, a computer, and software. I'd say the easiest software to use is the MakeMKV/DVD Audio Extractor combo. Use MakeMKV to backup the disc (defeating the copy protection), then use DVD Audio Extractor to rip the desired audio from the backup to your format of choice. Much simpler than an SACD.
I haven't ripped a Blu-ray in awhile, but I should, I still need to rip Chicago, Doobie Brothers and Yes, among others. I still remember it as being a bit of a pain with a number required steps, but like anything once back into it it shouldn't be too bad!
Ripping from SACD (once you get everything set up) is a breeze, follow the guide in the forum here. I use a cheap Sony player USB drive and my home network to rip straight to my computers hard drive, saved as .dsf files.
JediJoker
Audio Engineer/Enthusiast
And there's the rub. For the average user, it's much more complicated and likely requires a larger financial outlay (unless one already owns a compatible ripping device).
$50.00 for the player on eBay. No more (maybe less) complicated than Blu-ray ripping IMHO.
In Order of Preference ;
2 DVDA
2 DTS CDS
Single Bluray
2 DVDA
2 DTS CDS
Single Bluray
??? 24/88.2 PCM or ?
Yeah. That's pretty much what I do although I usually use XRecode3 rather than DAE. I find it takes much longer and with more effort than ripping an SACD with sacd_extract-gui. One click is all it takes.
OK. Then what's the 24x88.2 mean?
I don't know, that's what Foobar2000 reads the dsf files as, is it converting to PCM while it's playing the files?
Try the Merging Tech Hapi or Nadac although I do not believe they do better than my exaSound or Okto for playback.
You are the first one to state a (2nd) preference for DTS-CD's. When they first came out I thought (hoped) that they would be the rebirth of surround for music. Refreshing since the quad CD's that we were promised with the birth of the CD player never materialized. I felt that they provided great sound as well but when the non-lossy formats came out DTS didn't sound as good by comparison. I think that the main problem with DTS CD's is that some people would be bound to try to play it in a regular CD player and be blasted by the noise!
Similar threads
