Hello,
This is my very first post after only just finding HA and what an amazing forum this is...

Well, up until this year I've been a complete vinyl junkie and an audiophile (not sure whether to be proud or ashamed), but a recent housemove has seen all of my vinyl and a lot of my hi-fi kit go into storage until I find a bigger place to live, so before I moved I had the foresight to get all of my music onto my hard drive using EAC and LAME in Extreme mode.

Up until a couple of weeks ago I was using GBPVR so that my PC could be used as a jukebox and I could see what I was playing using the TV and control playback with my Hauppauge remote. This was fine for background music as I never thought the sound quality good enough to warrant serious listening.

Then of course I discovered Foobar and my listening habits were transformed. WOW! Having played around with the settings for a while, I can now quite safely say that it's the most I have enjoyed digital audio in a very long time, and quite possibly ever, and I've been through a good few high end CD players in my time.

In terms of sound output my PC is quite basic at the moment, I am just using the on board sound of my Biostar 200N cube which has AC97 sound. Up until I discovered Foobar I found the best sound I could get was to connect the optical out from the PC to my minidisc recorder and feed the line out from that into my hi-fi, but now Foobar has changed all that and the sound from the analogue out of the PC wipes the floor with the minidisc.

The hi-fi setup it's feeding into is a bit more special.. A Musical Fidelity A3.2 Pre-Power amp combination and a pair of Shahinian Arc speakers, with good quality interconnects throughout.

The settings I am using for Foobar at the moment are most of the recommended settings for good qulality sound - Strong ATH Noise Shaping, 32-bit output, real-time process priority (the PC isn't used for anything else while music is playing).

Sorry for the long post, I'm coming to the point now....

First question is this... In the re-sampling sections of Foobar it is said that upping the sample frequency makes "No perceived improvement" to the sound. I have spent the last few nights listening with upsampling on/off and different settings, and also comparing SSRC and PPHS re-sampling. In my listening tests I can quite reliably tell the difference between the two types, settling on SSRC for it's smoother sound and better projected vocals. I also find that upsampling DOES make an improvement and makes listening to digitally stored music more of a pleasure than I ever thought possible. The edginess in the sound I have always found with PCM has been removed and it gives much more insight into the recordings. I even find that 88200 is preferable to 96000. So I guess my question is.. Am I going nuts? Is this merely all psychological, have I been away from my vinyl too long and just forgotten how good it is, or does Foobar really transform what was once for me an 'inadequate medium' into something thoroughly enjoyable?

My second question is this... Obviously on board AC97 audio can't be anywhere near the perfect analogue output into my amps. I've been reading a bit about USB DAC's and I've seen a couple of kits available. I'm certainly up for spending some time with a soldering iron. Does anyone have any experience with these things and can they recommend it as a worthwile improvement?

My final question.. I have done some limited experimentation with the ASIO and Kernel Streaming plugins and so far have not seen any worthwhile benefits to using them over the waveoutput plugins, although I have read that a lot of people use them (although I think for different purposes). Is there anything really to be gained in terms of sound quality?

If you've read this far you've been very patient, and thanks in advance for any advice you can offer. As I mentioned, Foobar is a new discovery for me and It's the most exciting find I've had in a long time. It's as good as having had a new piece of hi-fi kit given to me for free!

Rich

Unless the sourcematerial was already at 32bit, this will only eat cpu-power without improving quality. The same goes for resampling. The only exception is if you have a soundcard which automatically resamples - various creative labs cards are most notorious for doing this. Thus, unless your soundcard sucks, neither upping the bitrate, nor resampling can improve soundquality. It is impossible.


Placebo-effect most probably.


This is true, but the reason which you imply is bullshit. AC97 cards automatically internally resample everything. This resampling sometimes can be done at suboptimal quality. In that case, it makes sense to let foobar do the resampling instead, so that the soundcard doesn't need to do it. This however, will ONLY work, if you resample exactly to the same samplerate/bitdepth to which your soundcard automatically resamples - else, you will have two lossy conversions, therefore getting worse quality than by doing no resampling at all.

To be clear hear - to get optimal soundquality, resampling should be avoided. But if the soundcard FORCES resampling anyways, then its better to do it at high quality, than at mediocre - thus, picking the least of two evils. The "udial"-testsample can be useful for testing various outputs until one gets perfect unmangled sound. Search for it. But be careful with the volume when playing this testsamples, else you may damage your ears and/or equipment.


No. All outputs are the same quality, unless your soundcard drivers suck. In such cases use either waveout or directsound, depending on which one works better.

As a sidenote - almost all of the issues you mentioned were already asked in other threads. By doing a search with keywords like "resampling ac97" or "kernel streaming" or "asio", you would have gotten the same answers without the need to spend that much effort in writing a long post.

- Lyx

Some interesting comments. Thanks for your reply. I have to admit I was expecting answers like this but I have a few comments.


I also thought this was more thank likely the case. So if we're not talking about an improvement in sound quality in objective terms, are we still talking about a *change* in sound quality in subjective terms which could be interpreted as an improvement by upsampling to 88000Hz. After all, there are numerous high end pieces of hi-fi kit on the market these days which use upsampling and I'm sure these are not _ALL_ marketing hype.


Probably not, not unless I have some sort of psychic ability anyway. I have housemates who like to tinker with the settings of Foobar and I've been able to tell within a few minutes of listening that the re-sampler settings have been changed, WITHOUT going anywhere near the computer. I should add at this point that my hearing is extremely good when it comes to spotting differences. I also spent 10 years in high end hi-fi retail, so that gives me at least a bit of qualification into being able to spot differences in playback. There is certainly a difference between the two to my ears and I find SSRC better.



What is bullshit? I didn't imply any reasons. I merely asked if anyone had had any experience with off board DAC's as a dedicated conversion and analogue stage outside of the PC case has to be better than the inbuilt AC97 audio. I completely understand about the internal resampling though and this makes perfect sense.

I will have a search for the "udial" testsamples and do a bit more experimentation. Thanks very much for the recommendation.


Yes, this is true, and I apologise for posting what has been posted before. As I mentioned however, this is a completely new forum to me and I'm sure you can appreciate that I would have had to spend a lot of time wading through all of the posts in order to get even some of the answers I was looking for. It was also useful for me as an introduction.

Thanks again
Rich

Please read item #8 in the HydrogenAudio Terms of Service.

If you don't want to "waste" the time for doing 20-30mins of homework, then why should other people waste their time giving you free support-service? You're asking for a gift, not for something you are entitled to - dont abuse it.

You may want to read this article.

- Lyx

Who said anything about not wanting to "waste" time? You strike me as someone who jumps to all of the wrong conclusions readily. There's no need to be constantly on the offensive. I find audio forums in general full of people like you and even with the best intentions I find your kind offputting.

I began to wonder whether I was in the right place after your last post, but I believe we're all after a common goal, and that's the pursuit of enjoyment of music, so we all have one thing in common. However, I completely understand the need to be objective to a large extent and I apologise for my unsubstantiated post. Consider me warned.

I'll come back when I've read the rest of the forums. No hard feelings ;-)
Rich

To claim what you hear regarding audio quality, you need to prove it by scientific methods, not just do a lot of things that sounds like improvements (those ASIO, kernel streaming things), enjoying the placebo feeling, with no real proof.

You need to educate yourself, learn the technology and science behind it (like resampling, and the reason behind).

You should do a listening test whenever you want to claim you hear an improvement in audio quality, and provide statistically confident results, especially in cases "if you are certain you hear a difference".

We have seen a lot of people like you here before, and some are less scientific and live in their own dreams. They did not listen to advices and led to flame wars.

That's why some of us here may sound harsh to newbies like you.

Then you should have no trouble following TOS#8 and posting the numbers confirming your results.

That said, I'll be quite surprised if you can. You'd certainly have a remarkably broken machine in that case.

Thank you all for your comments.

I have now seen the error of my ways and realised that I was far too eager to make a first post, making a complete fool of myself in the process. It's great that even though I feel I've had a good telling off, that it was for my own good and I can now begin a new approach to the whole subject of sound reproduction which actually means something.

One of the reasons I actually left hi-fi retail was because it was a dead end, there's only so much kit you can listen to and say "Yes, that's nice", I wanted to know more about why it was nice, which was why I spent 4 years after that at uni learning about sampling theory, DSP, transmission lines etc. etc. but obviously the old, subjective ways have not died away, so it's time to change all of that.

I've already spent some time with the ABX comparator and it's amazing how much you can *THINK* something is different, only to be proved completely wrong in a controlled environment. What a wake up call! It's taken me no more than a couple of days of being a member of these forums to realise that I've been living in a fantasy world.

Lyx - Please accept my apologies for wasting your time. I hope that in time I can prove of some use as I embark on a rather exciting voyage of discovery.

In all sincerity...
Rich