How do you play back lossy?

I’ve recently switchied to a resampling algorithm (Zoom Player with ffdshow audio filter) and am impressed with the difference in quality vs the standard gamut of players (WMP, WinAmp, etc). I’m new to resampling, and was curious what other people do to play back lossy-format files.

My setup: Run analog through the external inputs of my 5.1 receiver. As mentioned, I run ZPlayer w/ ffdshow installed and resample to 96kHz (Audigy 2 sound card; Creative outputs at 48kHz vs 44.1kHz).

I am curious as to why you are resampling the files at 96kHz then back to 48kHz?
It’s a waste of space - you will gain next to nothing by taking a 44.1kHz wav and converting it into 48.1kHz or 96kHz . All that will happen is that the program doing the conversion will attempt to interpolate new samples in-between the old ones but really it will sound pretty much the same as your original. I restore alot of music and record to the harddrive at 96kHz/32bit then resample back down to a standard wave file after processing it. But I do that inorder to get more info to work with from the initial file before processing.

I was reading an article (sorry I don’t have the link, I’ll look it up) and it indicated the higher resampling the better. Audigy 2 normally resamples to 48kHz, so I ran it first at 48kHz. After reading the article, I tried higher multipliers and it did sound better. From reading your reply it seems you’re thinking this is affecting the recording. I generally record at 44.1 as I’ve noticed anomalies in playback at higher bit rates (not so much on my home stereo, more when I burn CDs or playback on an mp3 player). If I had other FLAC players I guess I’d go lossless but everything I have aside from the computer is mp3/wma dependent. I can’t do bit perfect as my soundcard automatically scales to 48 kHz. I guess it could be a loss in cpu cycles, but it really does seem to help playback the higher I resample.

This is the thread I got the idea from, maybe I’m explaining poorly.
http://www.avsforum.com/avs-vb/showthread.php?t=528619&highlight=resampling

This is a quote from that thread:
“Upsampling relaxes the requirements on the analog reconstruction filter that follows the actual DAC circuit by performing much of the hard work in the digital domain, where it can be done more accurately and more cleanly.”

EDIT: In the same vein, I resize my DVD playback w/ ffdshow to 2.5x, which in turn is rescaled back to 1280x1024, the native res of my display. It has some remarkable effect, though a lot of that is due to the Denoise function of the filter.

Yeah what you read is wrong.

Your soundcard processes everything at 48kHz. Use 48kHz for the best quality. Any other sampling rate will be resampled to 48kHz (at a loss of sound quality). It is technically worse sound quality, regardless of whether it sounds “better to your ears”.

Higher sampling is only going to benefit you during the original recording from source to harddrive. After that you are just basically inserting 0’s to get to 96 or whatever,then depending on the software/filter(s) will interpolate from the surrounding information to get an approximation of what it thinks should be there. But if you like the way it sounds I guess thats all that matters.

I do understand what you are saying. First, this is the analog outs so it’s not just 1s and 0s. Even in the digital realm you’re looking at 24-bit representation by the soundcard, so a sequence of 24-0s and 1s would make up a “data point” or 2^24 possible combinations. It’d be a pretty worthless program to represent all 0s or all 1s. Let’s say I have 2 integers (representing the discrete sound byte), if the first value is 2 and the next value is 3, it’ll transition with an intermediary value of 2.5. You can apply whatever interpolative algorith to that be it linear, exponential, or even spline. Granted, that data is derived and can give rise to anomalies (that really come out when I try to equalize the data). No matter what, I will get resampling of the data as my sound card resamples the 44.1kHz files to 48kHz. What they’re saying in the thread is that you resample it digitally using the software then feed it back through the sound card rather than upsampling the analog signal with the sound card.

Now, I’m by no means an expert and perhaps the point is best left for someone with better understanding to defend. I’m sure I used wrong verbage, demonstrated a lack of understanding of basics, etc. and I do apologize. I suppose the best option is to upgrade the sound card and go for bit perfect. What I will say is that I’ve run the gamut of media players and the difference to me is comparable to the difference between analog and digital playback of CDs. It’s far richer than even the digital stream. Is that a consequence of a weak sound card? Maybe, like I said I’m just trying to find the best solution that I can.

Im not sure what you are saying about up sampling the analog sound. Once the sound is in the pc it is digital. Once it is in the pc I can upsample down sample or whatever. but you cant add any original sound that isn’t there when you record it to the harddrive. Upsampling is a two step process. taking the original signal up to the 48 or 96 kHz then interpolating to convert the O’s that the software adds to get to the higher rate to what it thinks should be there be it 0’s or 1’s. Like i said if you like it thats cool but all you are really doing is manipulating the original sound to something you like.
Personally I only use filters after recording to remove transient noise or harmonics, but I try and keep that to a minimum. If I want something a bit brighter during playback I will adjust the crossovers on the preamp or something but i try and leave the recorded sound as original as possible.

I meant its transmitted as analog to the receiver in my current set up. Of course you’re absolutely correct in all your points, but I guess without a better sound card I’m left with resampling. It’s definitely not a difference in brightness, more an increase in dynamic range. It brings out subtleties in the music and sharpens it, whereas I see brightness as increasing the highs and lows thereby creating an uneven sound that some people think is desirable. I’d say it’s much closer to the source material played across a toslink to my receiver, which is how I listen to CDs. I don’t have audiophile components by any stretch, just trying to get the most from my set up that I can.

EDIT: I did find the article I talked about here:


“The question of number of bits is another thing to consider. Does carrying extra bits increase the amount of information in our signal? Unfortunately, once we have sampled our signal, nothing can be done to increase the amount of information we have to work with. What carrying more bits does is that it prevents the loss of information. DSP algorithms and filters require additions, multiplications, and other math functions. If we are able to carry more bits in the results of these operations, we lose less information by chopping off fewer bits. Every truncation of a result will add noise to our signal. But now we can see that by balancing the number of bits we carry in our computations and by the amount we oversample, we can reduce the effect of this truncation in word length. One thing to note is that many products claim 24-bit word lengths, but yet only process internally at 20 bits.”

"There are basically two points of view regarding this upsampling an oversampling. The audio ‘purists’ want no additional processing on their signal and want whatever comes in from the source to come out as analog. They talk about zero oversampling DACs and such that are completely filter free both in the analog and digital domain. That is one extreme that some may argue is the purest since it avoids any digital artifacts and it’s quality relies on human perception by arguing that the human ear in itself acts as a brickwall filter after 20 kHz. Whenever we get into debates of human perception, the math and theory go out the window. Does it sound better without all the digital processing and filtering even with the image of the signal sitting just past fs/2? The energy past 22.05kHz is still present and you are still sending it to the speaker’s tweeter. How will the tweeter react to such out-of-band frequencies that are present? Furthermore, sending such a signal that is not limited in bandwidth could cause stability problems with wide-bandwidth amplifiers that have a high unity-gain crossing. The overall system’s signal-to-noise- ratio will be adversely affected as well. The DAC will also introduce frequency spurs all over the place. If we don’t filter them at all, what will their presence do to the sound? It’s a complicated problem and such a minimalist approach could introduce more non-linearities and negative effects, more so than the digital processing ever would. "

Here’s another which does support your stand more:
http://www.mlssa.com/pdf/Upsampling-theory-rev-2.pdf

“The sound quality of 44.1 kHz digital audio data can be dramatically improved by employing a “poor” oversampling digital anti-imaging filter having a slow roll-off in place of a “good” digital filter having a fast roll-off and a high stop band attenuation. It was shown that the ultrasonic images output by this “poor” filter is responsible for the improved sound quality, reducing certain forms of non-linear distortion such as that due to the differential non-linearity found in all DACs. There may very well be other, subtler, forms of non-linear distortion in DACs, which may also be reduced by signal-dependent ultrasonic dither.
In any case, there are certainly many other sources of non-linear distortion present in the signal chain. Some may question how such a small reduction in non-linear distortion due to differential non-linearity in DACs can be heard when much larger non-linear distortions are generated by loudspeakers, for example. The answer is that the non-linear distortions in question, like jitter-induced non-linearities, are uniquely digital in origin. Such digital distortions have no counterpart in the analog domain. It can be argued that human hearing is much more sensitive to certain digital forms of distortion as compared to the more common distortions of analog origin. For example, it is widely recognized that very low levels of jitter are audible even in the presence of much larger levels of harmonic distortion generated by loudspeakers.”

EDIT: I like what someone said in a thread concerning the original article I posted:
“I’d say the new generation of DACs (and almost everything else) is almost always cheaper but not necessarily better. Most advances in electronics seem more aimed at the bottom line than the high end.”

Interesting reading. Just a few points about what I saw in your quotes,
Quote:
Does carrying extra bits increase the amount of information in our signal? Unfortunately, once we have sampled our signal, nothing can be done to increase the amount of information we have to work with. Pretty self explanitory but I see what you are saying about trying to get the sound you like. Like I said if you like it thats all that matters.
And the comment about jitter is questionable in my opinion. It depends on the DAC etc… which leads me to the next quote.
Quote:
I’d say the new generation of DACs (and almost everything else) is almost always cheaper but not necessarily better. Most advances in electronics seem more aimed at the bottom line than the high end."

That could be true but anyone that is interested in finding quality audio devices either in the analog or digital realm can find them with a bit of research. I think price is a non issue. The sound card I use (Terratec DMX 6 Fire) has a much better sound than the Creative cards at a comparable price.

I really debated over the Terratec vs the Creative. I was a bigger gamer at the time, and I was already running a bit funky on my system (it took the hardware vendors a while to program where I’m stable now). I thought the drivers would be more stable with the creative standard. Hah! I couldn’t get the Terratec for less than $150 at the time, I got the Audigy 2 for $25. I can’t complain about the value of the Audigy at all, but now that I’m getting more into sound quality I’m recognizing the trade-off.

To tell you the truth I bought a Audigy Platinum without doing any research, I figured the Creative card was fine . I saw the pre-inputs which I liked I didn’t realise there was acard that had Phono inputs which would eliminate the need for a preamp for recording vinyl. After coming to CdFreaks I did some reading and someone here (minix i believe) pointed my to the Terratec and I ended up selling the Creative card to my brother.
Here is a little comparison data.
The Audigy Platinum not only has 32 bit processing but also includes an IEEE-1394 port and a new version of the EAX - the EAX Advanced HD. But be warned, the Audigy’s ability to support 24 bit/ 96 kHz sound is very relative. Unlike the DMX 6Fire 24/96, the Audigy Platinum is not a “true” 24/96 and cannot play or record a file of this quality, nor even work on it. Actually, the only task it does in 96 kHz is a linkup with another device via the S/PDIF input. The card and rack components are all supposed to be 24 bit/ 96 kHz compatible, but, in practice, the card is limited to 16 bits/ 48 kHz. Analog recording and restitution is possible in 24 bits/ 48 kHz, but, in fact, the processor downsamples in 16 bits and then upsamples. The card also has to be used with the sample rate conversion (SRC) software provided in order to work in 44.1 kHz. It’s just a pity that Creative has not given this product an automatic SRC hardware option by using, say, a Cirrus Logic CS8420 chip. And it’s really a pity that the card is not completely 24 bit/ 96 kHz compatible. But the Audigy can model, process and position several sound sources separately and in real time.
The DMX 6Fire 24/96 evolved from the EWX 24/96, and is based on an ICE1712-BBABA revision of an Envy 24 DSP by IC Ensemble. The card we tested is a 1.2C revision, i.e., the final retail version of the card as you will find in stores. Unlike the Creative Labs Audigy, the Envy 24 is a fully 24 bit/ 96 kHz compatible DSP. It supports 2 x 12 simultaneous digital or analog inputs/ outputs. This means the DSP manages 12 mono inputs and 12 mono outputs. Note that only 10 of the 12 inputs or outputs are 24 bit/ 96 kHz-compatible. The chip includes a 36 bit hardware digital mixer to ensure that none of the 24 bit channels loses in dynamic frequency.
You can guess which graph belongs to which card. :slight_smile: :slight_smile: :slight_smile:



Whoa - that top graph doesn’t look good! - It looks quite extraordinary … reminds me of some generic soundcard :slight_smile:

Wow! Talk about some crappy filter implementation. I really did make a bad decision, thanks for reenforcing that :stuck_out_tongue: (j/k). It’s also 5x scale, if they had a common scale it’d just be more dramatic. Is that jitter causing all of the fluctuations? I suppose that’s why resampling has such a strong influence on playback with my system. I should probably just buy a $20 chaintech and work towards bit perfect. Thanks for the input.

I believe it is the different chip in the Terratec. I usually dont read Toms hardware but he did a pretty nice comparison, also guru3d both are below. If you would like to read them


http://content.guru3d.com/article/sound/30/1

Tom’s is kinda hit or miss, but I do like their round ups. I bought my current case from a round up of theirs. Surprisingly, I’ve seen a lot of people recommending the X-Fi. After all the Creative Driver issues I’m reluctant. The current front runner at my budget would be the HDA Digital X-Mystique 7.1 for obvious reasons (DD encoding, been around long enough to work through the driverbugs). Unlike when I bought the Audigy, it seems like there are a lot of audiophile solutions. C-Media is coming out with some new cards just in time for the after Christmas market: http://www.bluegears.com/xplosion.html
It’s nice that the market is recognizing that computers aren’t all games.

This M-Audio seems to be a nice card for the money.$99.00 or so
http://www.very-clever.com/information/xbddddoxdxdhd

The Revo seems to have fallen out of favor with the audiophile crowd with all the new cards coming out. I’ve started the long process of re-ripping my audio to WMA lossless with EAC, it doesn’t even support resampling (which is OK, assuming I get another sound card). I’m glad they have a lossless format that supports all the various programs and devices, but suddenly a 1GB mp3 player is reduced to ca 40 songs. Passing through a Toslink with a DD signal should help in a lot of ways. From what I’ve read bit perfect runs on the M-Mystique, and it’s about the same price as the Revo.


EDIT:
I think I’ll wait for the XPlosion:
http://www.avsforum.com/avs-vb/showthread.php?t=605522