|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
In Reply to: "the softened, edgeless, forgiving, compressed, sand-brushed, fake cymbal, sound out of SACD." posted by Methos on April 25, 2004 at 20:31:11:
nt
Follow Ups:
You might recall, DG did listening comparisons where they couldn't reliably distinguish the DSD recording from 192/24 PCM. For you to claim that the final product is so utterly inferior to DVD-A...well, something just doesn't add up. :-)
DG was just being 'kind' to SACD by saying that. (It could have been much, much, worse!) :-)
...we'd probably get all sorts of entertaining hyberbole from the two of you, a bit like the IAR80 website. Then, if they made the tests blind...you'd be scratching your heads like monkeys (again). You might even prefer the DSD encoding! :-)
even how bad it can sound...but then again i wouldnt expect you to dig into it :-)
Or even tested?
I have recorded the SFSO Mahler 1st Sacd with 24/96 pcm.Noise floor sits at -75dB according to adobe audition analyzer scale.
Low level detail sounds grainy and noisy.The opening part of this recording is very suitable to crank up the gain and listen for sacd short commings.
it was on whole wheat bread. Microscopic particles of wheat were expelled. The grain you hear was actually in the air of the hall.
it's a live performance - if you crank up the volume you can clearly hear it's audience noise and reverb. i can tell it's reverb even at normal listening levels - that's what makes it a good recording.as for the low level detail sounding grainy and noisy, what can i say, your PCM recording is obviously not very good :-)
I did the same for a quite pcm recording.
Dead quit and far cleaner low level sounds.I even got -82dB from a very quite vinyl track with the same setup.
Of coarse you can hear audience noise and reverb. I even get the sounds of turning papers. The point is that these very low level sounds where grainy and noisy.
Better AD conversion is on it's way. (Motu 896HD)
it's very easy to get dead quiet for a studio recording - much harder for a live recording. and i wouldn't trust a live recording that sounds dead quiet - all the life has been processed away!i'm picking up -60 to -70 in my own living room, and drops down to -50 when the fridge turns on.
my point was the low level sounds on the mahler 1st were not grainy and noisy on my system (in fact, they were highly detailed), so the fault does not lie with the recording or the format.
I'm talking about how the quality of these very low level sounds heard through a headphone.
The gain was cranked op about 40dB. (That is gain before line level/headphone amplification.)Another disc where you can listen for low level detail is the Sony SBM demonstration disc. (Redbook listen to the water and chimes demo with and without SBM)
The sound quality of the demo track with the gain cranked up sounds far more natural than the low level detail on the Mahler sacd.
it's very difficult to make a "grainy and noisy" recording sound clean and detailed.on the other hand, it's very easy to corrupt a good recording into sounding noisy and grainy.
so tell me again why you think it's "included in the format"?
well, you could say perhaps your ears are more "golden" than mine, but let's not go there, shall we?
I listened through a magnifying glass. :)Through most domestic systems you are not capable to hear that kind of low level detail.
I hope you're not falling for the sacd has superiour resolution marketing crap. Because it really isn't true in it's 64*fs incarnation.
.
Christine,What Frank is reporting, ie amplifying the quiet section by 40dB could be a very long shot for a lot of software and converters. I wonder about some kind of post processing needed (eg whether any electrical artifacts or DC offset appear in the process). I would also check for the results at various bit rates.
But still, it's an interesting experiment, especially if you can compare it with an equivalent PCM version of the same material.
I plan to try the same thing to see if I get the same results, hopefully I can find a quiet section in one of my titles (don't feel like buying a Mahler title just to do a test :)Best
Eric
to keep the ad converters noise and non linearity out of the equasion as much as possible.
So what did you do exactly? You set your player to play a specific section, amplify it by 40dB (or just below clipping) and record the output from your receiver or preamp into your soundcad?Are you not recording / measuring distortion from the receiver if you do that?
Yes, you are including all the analog noise and distortion too if you capture from the player.With the Lynx you should be able to capture low sound levels with a good microphone and get raw pcm audio directly.
I used a 24/96 external usb interface M-Audio I had on loan.
I'm thinking of buying a Motu 896HD.
When I do I will put together a DVD Audio sampler with various tracks as example for various technologies. (SBM HDCD DVDA CD sacd)I also found a way to demonstrate the effect of dither.
I downrezzed a PCM recording with a bit crusher plugin to the point where bit resolution is 4 bits. By mixing in white noise through another channel before it is bitcrushed you can easily demonstrate the effect of adding dither noise.
Hey Frank,No wonder you've been so quiet :)
If you publish your tests, you should post it on this board...
With the Lynx, I was thinking of capturing through balanced inputs directly from my preamp, why would a microphone be better? (don't have mic preamps, either).
About the Motu 896HD: it looks cool, with so many inputs, I guess you can do multichannel recording and all kinds of cool stuff... but I heard their technical service is not always reliable so be careful where you buy it.
BTW, did you double-check your USB IRQs? I had some problems with that issue, and it clearly affected CE Pro (Adobe Audition) recording and playing at high resolution rates. I found that reserving one IRQ to the recording device helps a lot.
Best
Eric
You can use a microphone to capture low level sounds to test the low level sound quality of your dac. You can record familiar sounds to check if it sounds natural if you record with low gain setting.If you capture the output of a sacd or DVD Audio player you need a device with sufficient gain.
I didn't experienced problems with the usb(2) device.
do you mean: taking a well recorded SACD and recording onto 96/24 PCM (thus negating any benefit of DSD in the process), complain about the ambience and use it as evidence of DSD's "shortcomings", then compare it with a studio PCM recording as evidence of PCM superiority, then complain about the ambience being "noisy and grainy"?no thanks :-) it's already been done once, and that's enough.
Hold your horses, Milady...I think the experiment is interesting, especially if it's conducted with several digital recording devices. Of course the experiment is not scientific, but I have no pre-conception at all about DSD and will not draw any cheap conclusion from the experiment.
In fact, the experiment will tell very little about DSD, only about PCM. Since we don't know what happens with equivalent PCM data, the fact that it's DSD is irrelevant, and in fact, all the experiment says so far is that 24/96 PCM sounds "grainy" when amplified by 40dB on specific hardware :)
It could very well be that someone else, with different equipment and settings, would not find the result to be "grainy", or that the equivalent PCM material will also sound "grainy". Or it may be that with specific equipment and settings neither DSD nor PCM data captured at 24/96 sounds grainy.
What's the point? Well, from a practical point of view, since PCM is the only option available to me for recording analog data in high resolution, I find it interesting, because it may impact the way I record my SACDs at 24/96.
My first (very unscientific :) attempts at recording SACDs at 24/96 yielded similar results, I thought the sound was better at 24/48... but I also noticed that I had better results when recording a mono signal (less data to process) so the problem was on the processing side. So I upgraded my Motherboard to benefit from a faster bus, bought new hard disks, and changed the IRQ settings to avoid any conflict with my soundcard, and the results have greatly improved. But I didn't go as far as Frank, amplifying some sections by 40dB to check the sound...
I'll try that over the weekend, hopefully it will rain :)
Best
Eric
Frank has an ax to grind and can always be counted on to find problems with SACD. He is like some of our Supreme Court justices in the U.S., before the case is presented, everyone already knows what their ruling will be. It is always more interesting to hear tests from someone who is objective.
You are a bad judge of character.And already down to the point where you need to snipe at a person instead of contributing to a discussion.
There are no DSD benefits...If you do this with 24bit 192kHz the only thing you are listening for is the low level sound quality from each format. As long as you observe a difference in this comparison the results are valid.
Frank
isn't this a bit of circular logic?you don't believe there are any DSD benefits.
to prove your point, you record DSD onto PCM. so that even if there were any benefits, they would be lost.
then you hear what you think is a problem.
therefore, it proves your hypothesis that there is no benefit to DSD.
So you amplify the most quiet moments to hear artifacts or defects in the DSD version?Have you made a similar comparison with a dual format title (Diana Krall, Marvin Gaye, Swing Live, etc) ?
Best
It's not easy to get a quite part on these discs.
Mahlers 1st is well suited for this.NB it peaked at -6dB halfway through the first part. I adjusted the recording level just below digital clip level on that peak to get the best possible dynamic range.
Frank
I'll try to find some quiet or "silent" moments in my titles to see if I get the same thing you mention.Will post the results
Best
This post is made possible by the generous support of people like you and our sponsors: