advertisement


MDAC First Listen (part 00110011)

Status
Not open for further replies.
stephengrenfell said:
Dare I say it.. but does that include the SoC or does it need an additional rail. Or do we assume no SOC.

Do you plan to provide the external PSU, or should we try to source / provide that ?. Meaning would it makes thing easier for you and maybe help reduce your MDAC workload - just a thought.
Click to expand...

Does not include the SOC power requirements.

Thank you for the offer, but I'll need to source solution for the MDAC2 PSU - we have some ideas...
 
Coll said:
Would be nice to have some feedback to my post
Click to expand...
Some cds are now mastered in a way that the uadiophile market does not like (largely down to dynamic range compression). It seems that at least some people like this though.

It doesn't affect most of the music I listen to

FWIW I personally do not buy into the idea that vinyl is the gold standard or that there is anything wrong with 16/44 pcm. Perhaps vinyl replay has improved but there are huge flaws with the medium as whole: for example the hole in the middle of the record (pretty much all records) being off centre creates timing inaccuracies hugely greater than the tiny errors arising from jitter or anti/alias/ anti imaging filters. Even test records cant avoid this which is why wow and flutter figures are often given with a low pass filter. This may be fair enough when comparing turntables, but not IMHO when considering the medium itself
 
Julf said:
Yes and no. In general, digital recording, mixing and mastering technology has gotten much better. Unfortunately studio engineering skills haven't - or, rather, record labels push the engineers for a sound that is very "loud" all the time, requiring excessive dynamic compression.

Vinyl records a) can't mechanically take the excessive compression and b) tend to go for more "audiophile" oriented performances.
Click to expand...

I agree, so many recordings are spoilt, even completely ruined, by stupidly excessive compression. RHCP's Californication and Metallica's Death Magnetic are infamous examples.

I enjoy the composition of some popular music (Iggy Azalea, David Guetta), but I can't listen to it because of the horrible distortion on the vocals. Such a shame. It can be heard quite clearly on my setup even when listening to the Spotify and YouTube versions - for example Iggy Azalea's track "Black Widow".

And vinyl isn't always any better because it is often (not always) mastered to a similar level. When the vinyl is mastered correctly and the CD is not, then the vinyl will sound better because of the mastering not because it is on vinyl. The vinyl may or may not sound better for other reasons, however (depending on your point of view).
 
ChrisPa said:
So - again if PCM to DSD sounds superior - this points to problems with the effects of the modulator and how it is subsequently handled?
Because: PCM -> DSD should not inherently sound superior. There is no additional information and any "destruction"of the signal had already been done at time of recording/encoding

So any "improvements" are as a result of the (D to A) conversion process or artefacts of the conversion process
Click to expand...
upsampling to dsd should take the dacs anti imaging filter out of the equation and maybe the modulators, but if the dac is not converting in exactly the same format who knows what's going on? It might end up being converted back to pcm or remodulated.
 
mrflibble said:
And vinyl isn't always any better because it is often (not always) mastered to a similar level.
Click to expand...

Except vinyl can't actually physically take the same kind of ultra-loud, ultra-compressed mastering we see on many CDs.
 
JohnW said:
For sure thats not what I've heard in the past from Vinyl - even with my systems deficiency the Analogue source is far more "real", far far better sound stage and maybe more importantly depth... every instrument & voice has its own local in space, and no matter how loud the other instruments are you can clearly hear each instrument if you focus your attention to it - it put me into somewhat of a depression for a few days after my first listening session when it was forced home yet again how we lost SOOOOOO much going digital IMO :(
Click to expand...

Whilst it doesn't fix the problem of the decimation filters in the delta-sigma ADC used during the recording process, the use of a Non-oversampling DAC or a multibit DAC should preserve the samples present in the digital signal. These types of DACs tend to sound more analogue like. For those interested, it may be worth trying a cheap NOS or multibit DAC to hear if it is to their taste.

There is a chap in France who makes little TDA1543 NOS DACs and sells them on ebay, searching for "nos tda1543" on ebay will bring them up. They are a bit lacking in resolution but are easy to listen to.
 
Julf said:
Except vinyl can't actually physically take the same kind of ultra-loud, ultra-compressed mastering we see on many CDs.
Click to expand...

Yes, you are right. But I believe the vinyl of Death Magnetic was still pretty bad?

If the original recording is performed digitally with an ADC, then there will a further conversion back to analogue to record the signal onto vinyl. The vinyl would likely exhibit many of the issues that the CD has - the difference being the digital to analogue conversion has happened at the production phase rather than the playback phase. Of course, in the good old days, I believe master recordings were made to analogue tape and the vinyl was cut from these.
 
mrflibble said:
If the original recording is performed digitally with an ADC, then there will a further conversion back to analogue to record the signal onto vinyl. The vinyl would likely exhibit many of the issues that the CD has - the difference being the digital to analogue conversion has happened at the production phase rather than the playback phase. Of course, in the good old days, I believe master recordings were made to analogue tape and the vinyl was cut from these.
Click to expand...

Indeed, but thos good old days are long gone - these days pretty much all recordings are recorded, mixed and mastered digitally.
 
Richard Kimber said:
Has this changed? I thought you said earlier that MDAC2 was DSD256. I wish there was a definitive list of MDAC2 specs available.

- Richard
Click to expand...

Richard,

We have DSD512 working on the bench - but no OS / Software can support such a data rate - I dont recall how we tested it.

Basically the hardware can support DSD512, but the PC software / OS will take time to catch up.
 
JohnW said:
Richard,

We have DSD512 working on the bench - but no OS / Software can support such a data rate - I dont recall how we tested it.

Basically the hardware can support DSD512, but the PC software / OS will take time to catch up.
Click to expand...
It's one more step up - DSD512 is commercially available, DSD1024 is possible in "lab conditions". Google gives me ie. http://www.diyinhk.com/shop/audio-k...to-i2sdsd-pcb.html#/xmos_option-xmos_768k_pcb amongst the other high-end DACs.

Driver wise, I don't see any rates hardcoded for native DSD on Linux, so it can theoretically support anything the (USB) device announces. That is, as long as it fits ie. 32bit signed integers, I assume some code uses those for Hz. Still, that gives you some ~2GHz sample rate to work with, though you might want to use USB 3.1 for that.

edit: Another cheap one - http://rover.ebay.com/rover/1/711-5...0001&campid=5338728743&icep_item=141316861730 ... somebody should buy one and figure out how it plugs into the linux/sound subsystem. :) (Might do it after xmas.) I remember looking at how they implemented native DSD and it was along the lines of defining it as a single-bit PCM or something like that. As the kernel just passes samples around, it presumably works fine.
Code: 
./uapi/sound/asound.h:#define   SNDRV_PCM_FORMAT_DSD_U8         ((__force snd_pcm_format_t) 48) /* DSD, 1-byte samples DSD (x8) */
./uapi/sound/asound.h:#define   SNDRV_PCM_FORMAT_DSD_U16_LE     ((__force snd_pcm_format_t) 49) /* DSD, 2-byte samples DSD (x16), little endian */
./uapi/sound/asound.h:#define   SNDRV_PCM_FORMAT_DSD_U32_LE     ((__force snd_pcm_format_t) 50) /* DSD, 4-byte samples DSD (x32), little endian */
./uapi/sound/asound.h:#define   SNDRV_PCM_FORMAT_DSD_U16_BE     ((__force snd_pcm_format_t) 51) /* DSD, 2-byte samples DSD (x16), big endian */
./uapi/sound/asound.h:#define   SNDRV_PCM_FORMAT_DSD_U32_BE     ((__force snd_pcm_format_t) 52) /* DSD, 4-byte samples DSD (x32), big endian */
Hmm, looks like it's just 2/4 byte blocks. Reminds me of DSF/DFF which really just contains blocks of DSD bits. :)
 
This site contains affiliate links for which pink fish media may be compensated.
JiriJ,

Its just my opinion, but I dont consider Linux a typical everyday users OS... its too specialised - you need to be a computer geek to use it :)

I'm getting too old to want to spend any more of my life fighting computers.. it was ok when I was 15 years old with nothing better to do with my life!

Jarek had DSD512 working here in the lab and hes the kind of guy that likes to mess with such detractions :)
 
Richard Kimber said:
The new Matrix Audio X-SABRE Pro claims to be capable of 1024

- Richard.
Click to expand...

Its not really a hardware issue - but the Host software, we have no means to test above DSD512 but when there is a solution then there's no reason we cannot support it.

I'm not aware of any "true" DSD1024 capable ADC's so at the moment the feature is rather moot.
 
JohnW said:
Its just my opinion, but I dont consider Linux a typical everyday users OS... its too specialised - you need to be a computer geek to use it :)

I'm getting too old to want to spend any more of my life fighting computers.. it was ok when I was 15 years old with nothing better to do with my life!
Click to expand...

+ 1 billion or more.
 
There are linux distros that are perfect for everyday users. It's just that the distros you want to pick as a base for a special embedded application are of course geared towards experts.
 
Status
Not open for further replies.


advertisement


Back
Top