Indeed, but not enough to make a desease of it.
Let say 24 bits are a requisite when we have to manipulate the contents, 16 bits a satisfying minimum when the levels are well normalized, IE consumer reproduction.
I would fight more for 96kHz sampling frequency.
For playbak, up/sampling or over sampling makes it possible to avoid/limit the filters needed after the D2A conversion.
you still face the problem that the above app 20KHz content is lost. while we cant hear a 20 KHz sine alone we can hear the super positions on 20 and 21 KHz as they will produce lower frequency content.
The higher frequency content is still present in a Vinyl pressing of a 24/96 studio recording, but lost with the CD issue.
you still face the problem that the above app 20KHz content is lost. while we cant hear a 20 KHz sine alone we can hear the super positions on 20 and 21 KHz as they will produce lower frequency content.
The higher frequency content is still present in a Vinyl pressing of a 24/96 studio recording, but lost with the CD issue.
Yikes, Joe's Filter isn't directly about the powersupply.
He loads the output of the SD DAC with a filter that shaves off a few dB around 20k. This apparently takes away some of the 'SD DAC harshness' even when the fewdB@20k is corrected for digitally (preDAC) or anally
postDAC .[The filter is manually dialed-in, and is apparently interacting with the internals/deepprocess of the SD DAC to properly load it, or compensate some phasey drifting, or something, I don't git ... ]
That's Joe's Filter.
http://www.diyaudio.com/forums/lounge/249418-dac-filtering-rasmussen-effect-56.html#post4219861
http://www.diyaudio.com/forums/lounge/249418-dac-filtering-rasmussen-effect-54.html#post4214471
PS Sounds like you understand the internals of a typical SD DAC, take a look ....
Last edited:
Sorry to press the question (you guys wander all over the place
Jan
Well a big limitation is backgound noise, if you live in a city, even at 2 am the traffic noise limits the bottom end, and then at the top end you have to consider the loudness versus time safe limits, occasional drum hit to 110dBA is not going to do much damage, whereas a sustained compressed, and clipping track at 100dBA is going to affect your hearing much sooner.
For folks lucky enough to live in the middle of nowhere then backround noise is low enough to make a greater range reachable.
Jan
My room is NC15! In reinforcing symphony orchestras live I have found 20 dB of headroom is not quite enough, I aim for 30 dB. Allowing for 30 dB critical band signal below noise capabilities that allows for a listening level of 73 dB. That level would have significant timbre shift.
The other issue is in order to get 118db from sixteen bits what else is really required.
My room is NC15! In reinforcing symphony orchestras live I have found 20 dB of headroom is not quite enough, I aim for 30 dB. Allowing for 30 dB critical band signal below noise capabilities that allows for a listening level of 73 dB. That level would have significant timbre shift.
The other issue is in order to get 118db from sixteen bits what else is really required.
Does the email give an IP address ?
http://www.diyaudio.com/forums/forum-problems/269545-someone-tried-log-under-my-name.html
The superposition of two tones only produces a difference tone if intermodulation mechanisms are present.
Sorry to press the question (you guys wander all over the place
Jan
Yes, of course.
My starting point was/is that if you download an uncompressed 24b/96KHz file and play it back, it sounds better than listening off a spinning CD of 16b/44.1.
As simple as that.
Then I said the dynamic range was larger and the distortion was lower using 24/96. And, at the level we listen, we are not able to quote the 0fs level's distortion as is often touted. That would only occure at peaks in the music....but we are in the middle of 8bits most of the time with a 16b system.
There have been a lot of work-arounds to get the distortion covered (dither) and compression moves more of the average level to higher bits. And, there are the lossy codecs and iPODs. It all works fairly well and I have been listening to CD since day one and doing my part to support the economy with their purchases.
However, there is No doubt that download files direct from master 24/96 is an improvement. I would say a major improvement.
Along the way, thinking minds can tell why analog has always sounded better..... near infinite resolution limited only by distortion and noise--- both of which are extremely low over the range of levels we listen. The only problem is how to get direct feed sources in analog -- from master tapes or a live broadcast (FM). Only the live broadcast is practical but all too rare.
Digital can be very close to analog IF we get the bits up and the sampling rate up higher. 24/96 is VERY good. And, yes, it will mostly be done in IC's.
THx for listening--RNMarsh
Last edited:
I hope you feel dirty adding sound reinforcement to something that shouldn't have it
. I went to a concert at the 02 in london a few weeks back and the reinforcement was so much that you couldn't hear a single instrument directly. Never going there again. I think the desk monkey was used to rock concerts. Looking at some of the best recordings I have for dynamic range they come out at DR19 with the foobar plugin, with RMS levels around -30dB. I am suspicious what the DRM plugin actually tells us, but at least I can compare. So in my (very) small room 70dB average level gives 100dB peaks at the sofa.
Actually pretty realistic for the levels for the listening positions I can afford in most venues. My happiest musical times are promenading in the albert hall. the classical equivalent of the mosh pit, in a place with perhaps the worst accoustics you can get, but the experience is unlike anything else and only £5 a ticket.
Dick if you think I've been advocating 16 rather than 24 bits you are sorely mistaken. I use 24/96 for everything but realize that careful production down to 44.1/16 is a lot less noticeable (or un-noticeable) than some claim.
As jcx asks where are the industry sponsored studies? What has been offered is to me no more compelling than endorsements for Shatki Stones.
That Benchmark probably uses the ESS "32bit" DAC's. 32 bits purely in the minds of the marketeers. To be fair, as good an implementation of Sigma Delta techniques as is out there.
Yes, no actual details - because how does one describe hearing distortion, and making a mental jump to the right cause of that - at best it would be extremely involved ... and every situation is different.
Sorry about that ...
I don't tolerate distortion but i don't listen at very high volume and my system is quite respectable so i don't hear any unless the recording contains it. Some rock bands deliberately include distortion in their material so seeking to remove that would produce a false outcome.
Maybe i should have asked if you could give us some examples of what you found wrong in a particular setup and how you rectified it.
I found that well designed circuit layout is crucial, then decent components with closer tolerances can provide some improvement. Recently i moved my power amps close to my speakers so i could use short speaker cables, my buffer (DCB1) is easily able to drive 15 foot interconnects, that brought some benefits too.
As for cars, i used to be a mechanic back in the 1980's before i became a marine engineer. Believe me i've heard some crackers. It's making a 'going round' noise etc. A test drive usually helped, also i had a quite sophisticated Crypton engine tuner at my disposal which i was trained to use. Cause and effect Frank, just like in audio.
One more thing, if it ain't broke, don't fix it
Dick, just picking on this one: A phono system with 60dB DR has of course only 60dB resolution, whereas a digital system with 110dB DR has about 50dB MORE resolution. Which one gets closer to infinity?
So if you prefer that phono system (and often I do too) it's NOT because of the higher resolution (it has much lower), but just because you/I like it better. Not much to argue about that. But don't try to tell us it is BETTER* - it isn't, not by any stretch of the imagination.
It's all relative (at least since 1916 or so).
Jan
*assuming that by better you mean more accurate.
I have a number of HD tracks 96 and 192 KHz 24 bit albums.
I play them back on the same set of ESL speakers and subs through 2 sets of electronics. The first is a very new home theatre reciever set to 2 channel mode with 100 watts a channel and playing back a 192 kHz file.
The entire second electronics chain is all discrete jfet/fet DC servo DAC, jfet/fet preamp, and, class A 100 Fet watt power amp. Low level electronics are powered by shunt regulators. Same DAC chip as the reciever.
The same files played back through the reciever in straight through mode sound dull, undynamic, lacking detail, and mid-Fi at best.
Switching to the all discrete signal chain brings a dramatic improvement in dynamics, resolution, smoothness, bass impact, image depth - you want to listen longer.
I agree with Richard - higher bit and sample rates are clearly audible - though to a lesser degree on the reciever. The playback chain matters, little point in higher bit and sampling rates played back on cost optimized, feature laden, mid-fi electronics. The investment, time and monetary, in building or buying better electronics and speakers shows as a more realistic, involving performance and higher bit and sampling rates make a better system's qualities even more obvious. Involving is the key word, you feel closer to the original performance.
I play them back on the same set of ESL speakers and subs through 2 sets of electronics. The first is a very new home theatre reciever set to 2 channel mode with 100 watts a channel and playing back a 192 kHz file.
The entire second electronics chain is all discrete jfet/fet DC servo DAC, jfet/fet preamp, and, class A 100 Fet watt power amp. Low level electronics are powered by shunt regulators. Same DAC chip as the reciever.
The same files played back through the reciever in straight through mode sound dull, undynamic, lacking detail, and mid-Fi at best.
Switching to the all discrete signal chain brings a dramatic improvement in dynamics, resolution, smoothness, bass impact, image depth - you want to listen longer.
I agree with Richard - higher bit and sample rates are clearly audible - though to a lesser degree on the reciever. The playback chain matters, little point in higher bit and sampling rates played back on cost optimized, feature laden, mid-fi electronics. The investment, time and monetary, in building or buying better electronics and speakers shows as a more realistic, involving performance and higher bit and sampling rates make a better system's qualities even more obvious. Involving is the key word, you feel closer to the original performance.
Last edited:
How have you compared tho? Have you resampled some of your recordings down to 16/48 or HD re-issues of things you already have? Or can you innately tell a 16/44.1 recording just by listening?
The quality of the down sampling process is extremely important. See for instance the SRC comparison site and the literature on shaped dither. a fair comparison would be with files generated by someone capable of driving the software and the knowledge to verify the results. Not someone using what "everyone thinks is pretty good".
- Status
- Not open for further replies.