Thread: bitstream pure DSD, my 2 cents

Posts: 74
Page: prev 1 2 3 4 5 6 7 8 next

Post by Iain February 2, 2012 (31 of 74)
Yoropiko1 said:

Agreed, the spacial aspect of the sound in particular is what seperates the sound above all else in my opinion too, if I close my eyes there is definately more dimentionality to the stereo mixes ( although I must admit the differences in multi channel are very difficult to seperate if I'm honest ). I can only put ths down to the 176.4Khz conversion retaining more of the original information derived from the DSD. I can liken it to the difference between CD and high bit rate ( 320kb+/sec ) MP3, the music still basically sounds the same but the MP3 has that something missing and is less enjoyable to listen to.

BTW, I'm forced to listen to SA-CD in 2-channel mode exclusively because of the inconsistent LFE issue. 90% of my SA-CD collection is 5.0 whilst the rest is 5.1. Hence my remark of spatial clarity.

I have small, but quality speakers all round.

I find that LFE issue quite odd, but my Denon AVR will add a .1 track to all 2-channel playback, Haven't found a configuration option on it for multi-channel yet.

Wish SA-CD OEM would get this inconsistency sorted.

Post by Yoropiko1 February 2, 2012 (32 of 74)
An interesting point. You would assume that an all in one box system like that should convert the DSD to analog but I dought it very much, simply because the system would also have to deal with PCM sources of various kinds from CD, Dolby and DTS. The price point and the quality of the rest of the systems componants would perhaps not justify the use of high spec DSD DACs. I openly admit I could be wrong of course as I am not familier with the AVD series. If you think of an all in one system as a budget seperates system all wired into one box, all sharing the same power supply, etc then I think even in the ulikely event a DSD>>>Analog conversion was possible there would be almost no chance you would hear the benefit. In fact despite Sony packing SACD playback into lots of its componants, such as its line of Blu ray players for example, very few of those purchased will ever see an SACD! : (

The phrase " amp modules use DSD directly " are open to interpretation. Consider my own Yamaha RX V2067 can accept a raw DSD feed which would make you believe it should convert to analog but it doesnt. My Receiver uses DSD directly too.........directly to the DAC to be converted into PCM!! LMAO ^ _ ^

Joking aside, kudos to you for getting an SACD capable system for such a bargain price, I'm certainly not knocking it and you have lost nothing if SACD's dont sound any better than regular CDs given all the other functions such a system can offer. If it does sound better then you have got yourself a cracking little system, just dont get addicted, listening to SACD can seriously ruin your bank account! ; )

Post by Kal Rubinson February 2, 2012 (33 of 74)
Iain said:

I find that LFE issue quite odd, but my Denon AVR will add a .1 track to all 2-channel playback, Haven't found a configuration option on it for multi-channel yet.

Generally, if the main L/R speakers are set to SMALL in the menu and the subwoofer is set to ON, the AVR will reroute bass to the sub, regardless of the number of channels. If your AVR isn't doing it, you probably are using DIRECT or some other setting that precludes bass management.

Post by Whitehall February 2, 2012 (34 of 74)
This document gives some description of Sony's method of using DSD within amplifiers:

http://www.docs.sony.com/release/ES_STR_05_Final.pdf

They seem to use the bitstream from DSD and use it to flip power tranistors into strong positive and negative pulses which are smoothened and filtered into an audio signal at the speaker terminals. They seem to make much of the lack of PCM within. In fact, their devices have ADCs to change analog inputs into DSD!

Not sure if this is pulse width modulation (PWM) or not but I strongly suspect that the AWD reciever I have uses the same basic chip set.

Post by Iain February 3, 2012 (35 of 74)
Kal Rubinson said:

Generally, if the main L/R speakers are set to SMALL in the menu and the subwoofer is set to ON, the AVR will reroute bass to the sub, regardless of the number of channels. If your AVR isn't doing it, you probably are using DIRECT or some other setting that precludes bass management.

Sorted......finally!! : )

Thanks for the information. Problem was an obscure set-up option that was causing the sub-woofer to disconnect on 5.0 media.

Post by steviev February 3, 2012 (36 of 74)
Yoropiko1 said:

Well I guess I have to take back my previous comment, because today I did in fact receive a reply to my enquiry to Yamaha UK, here is what they said;

"Thanks for your email and for your interest in Yamaha A/V & Hi-fi.

With regards to your question on your RX-V2067, the DSD signal gets down sampled
to 88.2kHz/24bit (straight mode) first and then transmitted to DAC for analogue conversion,
in case of enabling Pure Direct mode.

Hope this helps.

Many thanks

Yamaha Technical A/V

This is a very confusing answer from the Yamaha tech, and I'm pretty sure it's 50% incorrect.

I have a Yamaha RX-A1000, and it performs much different in "straight" mode versus Pure Direct mode.

The Yamaha tech says that DSD is down sampled to 88/24 in "(straight mode)". In Straight mode (which is an explicit setting for this receiver), the RX-A1000 performs distance correction and parametric EQ, which definitely requires conversion to PCM. In Pure Direct mode DSD, however, the receiver does absolutely no processing: no EQ, no distance correction, nothing. I have confirmed this by experimenting with extreme and absurd distance and EQ settings that make it obvious when the receiver is or is not altering these parameters.

However, if I set my SACD player to output PCM (176/24), then the A1000 does distance correction only but not parametric EQ, even in Pure Direct mode. This is true of any PCM output, including RBCD and Bluray.

Also, my receiver seems to perform worse when modulating pure DSD. During certain very loud passages on some SACDs, the A1000 will stop modulating at all: it will go instantly and utterly silent for the duration of the passage. And it's very consistent -- no matter how loud or quiet I set the receiver, it will go "blank" for that passage for the exact same stretch of time. Also, I have one SACD that "pops" annoyingly (and again, consistently, at the exact same moments every time) only in pure DSD, but these sonic artifacts disappear in PCM. Obviously, this is disturbing and so I keep my SACD player set to PCM at all times.

And like others who have A/B'd DSD versus 176/24 PCM, I can hear no difference, other than those disturbing digital hiccups and artifacts.

Post by Yoropiko1 February 3, 2012 (37 of 74)
steviev said:

This is a very confusing answer from the Yamaha tech, and I'm pretty sure it's 50% incorrect.

I have a Yamaha RX-A1000, and it performs much different in "straight" mode versus Pure Direct mode.

The Yamaha tech says that DSD is down sampled to 88/24 in "(straight mode)". In Straight mode (which is an explicit setting for this receiver), the RX-A1000 performs distance correction and parametric EQ, which definitely requires conversion to PCM. In Pure Direct mode DSD, however, the receiver does absolutely no processing: no EQ, no distance correction, nothing. I have confirmed this by experimenting with extreme and absurd distance and EQ settings that make it obvious when the receiver is or is not altering these parameters.

Interesting, perhaps the RX A1000 has different DACs to its Euro version the RX1067. I would assume the general construction and internal componants would be the same ( the only difference being the AC input, 110v USA, 240V Europe ) but who knows if maybe different regions get slightly different specs or why would they even bother to give them different names ( ie why not call all the receivers in the range RX A1000 whatever region it is sold?)
When I select " straight" mode on my Receiver and input stereo bitstreamed DSD I can see the L channel, R channel and SW icons lit up on the front display. Stereo SACD's would not have a dedicated Subwoofer channel so I assume when I see this my reciever is converting the DSD to PCM and adding a SW channel artificially ( which it would have to do to enable any Bass management for example ). However if I then switch the " Pure Direct " on, the SW icon dissappears, leaving only the L and R indicators on the display. So is this processing the DSD as is without PCM intervention?
Having said that if I pop a PCM RBCD into the PS3 and send it via HDMI to my receiver in straight mode, I again see the L , R , and the SW channels lit up. Of course the source is PCM anyway, but my receiver takes it upon itself to create a SW channel from a strictly stereo CD which of course would never have a SW channel of its own. If I switch into " Pure Direct " the SW icon vanishes and just the L and R channels appear.
So needless to say I am left very confused indeed, thus I can only really go by the two things that point to my receiver converting DSD into PCM, that of the DACs inside the RX V2067 and of course the answer I got from Yamaha.

Until somebody can conclusively confirm that my RX V2067 can in fact convert DSD to Analog I will have to assume it doesnt : (

Of course the only other questions that still baffle me is if my Yamaha has PCM 1789 Dacs that are NOT DSD capable, how does it even accept the incoming DSD signal in the first place to convert it to PCM ( does anybody know?), and why on earth would Yamaha even make an amp able to accept raw DSD then scrimp a few dollars to fit it with DACs which cannot do DSD>>>Analog when there are alternatives ( which are fitted to the RX V3067 ) which would likely add a very small amount to the cost of the product??? Looking at TI own site suggests a DSD1796 DAC costs less than $4.00 and the PCM1789 costs about half that ( I accept these prices are based on bulk orders of course ), hardly a massive difference which could have been easily absorbed into the selling price. Makes me wonder if most audio products these days are designed by music entusiasts or accountants!!

Post by tailspn February 3, 2012 (38 of 74)
Yoropiko1 said:

Of course the only other questions that still baffle me is if my Yamaha has PCM 1789 Dacs that are NOT DSD capable, how does it even accept the incoming DSD signal in the first place to convert it to PCM

The DAC IC is the last in a line of chips an any player of processor. Preceding the DAC chip are formatter and digital signal processing integrated circuits that provide the feature set of that unit. The "why don't they" question is a marketing, not a technological question.

Post by Yoropiko1 February 4, 2012 (39 of 74)
tailspn said:

The DAC IC is the last in a line of chips an any player of processor. Preceding the DAC chip are formatter and digital signal processing integrated circuits that provide the feature set of that unit. The "why don't they" question is a marketing, not a technological question.

Thanks for that, so am I to assume a DSD stream would bypass the formatter and DSP circuit if a DSD DAC were used? I ask this because if there is non DSD capable DAC, converting to PCM also introduces yet more conversion of the DSD signal before it even gets converted to PCM by the DAC.
That cant do the original DSD signal much good : (

Post by Yoropiko1 March 6, 2012 (40 of 74)
steviev said:

This is a very confusing answer from the Yamaha tech, and I'm pretty sure it's 50% incorrect.

I have a Yamaha RX-A1000, and it performs much different in "straight" mode versus Pure Direct mode.

Futher to my investigations re Yamaha's responce to my Receiver the RX V2067 converting all DSD to PCM at 24Bit 88.2Khz, I decided to obtain a copy of the review of the RX V567 receiver, a lower end model from the same "7" series as tested by Hi Fi World September 2010. Now before I go any futher, I should point out that I do not know which DACs were fitted to the 567, even the review only describes the dacs fitted to this receiver as Burr Brown 24Bit/192Khz ones and doesnt state the chip type used ( ie if its a PCM or DSD model ), but it should be reasonable to assume that an inferior or at best identical DAC would be supplied with the 567 compared to the more expensive 2067. Certainly the 2067 does not possess DSD Dacs, it is fitted with Burr Brown PCM1789 which I believe are not DSD compatible.

The review in question is somewhat ambuguious in my opinion but does suggest that the 567 does convert DSD to analog without a PCM conversion. Hi Fi World are well respected in the industry and offer detailed Measured Performance graphs and the like to accompany their subjective reviews.

The first quote from the review is as follows;

"The biggest surprise came when, with the Cambridge 650BD set to output DSD, the Yamaha recognised it as such and played it, to provide high resolution surround sound from SACD discs. Measurement showed DSD decode was very linear too, distortion from SACD being lower than that from 24Bit PCM - impressive for such an unpretentious budget receiver."

The second quote from the review is as follows;

"Spinning Lang Lang playing "Rhapsody of Paganini" made me smile as DSD lit on the front panel display: not bad for a £380 receiver, I thought."

Now the above two quotes suggest only that the Yamaha accepts DSD, ie the front panel indicated it as such, which of course the RX V2067 also does on its display and On Screen Display as well. As we all know, simply accepting a DSD stream doesnt in itself mean that the receiver converts that stream into Analog without a PCM conversion so the reviewer might have been under the impression that DSD was being converted to analog simply because the receiver was accepting the raw DSD bitstream. Thus the review has been incorrectly interpreted by those who read it beleiving as indeed the reviewer has suggested that the 567 is converting DSD to Analog when it doesnt explicitly state this.

However.........Look at the Measured Performance section, this suggests otherwise and in fact does state that there is a possibility of a DSD >>> Analog conversion given the distortion figures. Again the quote from the test results is as follows;

"SACD ( DSD code ) reached 28Khz and distortion very low, 0.16% at -60dB. As this was better than 24Bit PCM it looks as if this receiver has proper DSD convertors and does not transcode to PCM."

Distortion ( -60dB, 16/24Bit/DSD )
0.28, 0.22%, 0.16%

Now as I've stated above, the results of the 567 do not mean that the 2067 would be the same or even for that matter have the same internal componants, but the two models are part of the same family, released at the same time, it would be most unusual for a lower end model to feature something which models above it do not have, although of course that is not impossible. It could be that the 567 does feature DSD capable DACs but I dought it at this price point if models above with the exception of the top model the 3067 dont have them.

Still as it stands I have to accept Yamaha's word regards DSD conversion to PCM until somebody out there can without any dought convince me that my own 2067 can do a DSD >>> Analog. Still it makes interesting reading and I am glad I have had chance now to read the review myself in full, having seen it quoted several times by several others on this and other forums. I hope one day manufacturers will be more open about the features suppoerted by their products, sadly it seems there is a great deal of confusion all round.

Page: prev 1 2 3 4 5 6 7 8 next

Closed