XXHighEnd

Ultimate Audio Playback => Interesting Music / Testmaterial => Topic started by: Flecko on December 09, 2010, 11:23:14 pm



Title: Creating a common test library
Post by: Flecko on December 09, 2010, 11:23:14 pm
The changing results from my setup in the thread "SPDIF Cable, longer is better! (Interesting articel)" http://www.phasure.com/index.php?topic=1425.0 made me mistrust myself and I asked myself "WHY?" ;) Ok, everyone knows this, one day you in that mood and next day in a different. So, I am to blame :) But are there systematic errors that could have been avoided? Yes there are. The quality of the record should be of a quality, that is not just very good but exceptional. There should be a setup of songs that garanties that every song you pick for the test has this quality. If the record itself contains just a slight hereable distortion it might sound better with the wrong set up. I never realy tested and compared the recording quality of my CD's. I had an opinion on the soundquality but I never realy tryed to find the best sounding records from my library. So I used to test with some good records but not consistent with exceptional records. And this was wrong. Hence, I took the files from which I thought they might be very good and copied them to a folder. I choose different music styles to have a variation of sounds and instruments. Then I compared them always looking for minimum distortion from recording and best mastering. Now I have a list of tracks I will provide you here. The idea is that I will share this tracks with friends from this forum, that are interested to build a common library. As you have receifed this tracks you will listen to them and you could suggest yourself some tracks that you think are of exceptional quality. Then, if some people will take part, we will have a set of very good records and then we can vote which should be in the test library and which not. Next time if we discuss soundquality of a special setting we can reffer to track XY. So we can talk aout the same things and not comparing aples and oranges. And I am sure we all will learn a lot from each other about interesting music. It has some "illegal" touch but I am fine. I own all records I share with you. We will not make them available to everyone, so this has to be done over PM. There we can provide the links. I think this should be ok (Peter are you fine with this? I wont send any files until you give your ok for this). You post your list of songs in this thread and then you can be contacted by PM and you will provide the download address. Of course only to people we know. After we decided which songs should be in the test library, the not needed tracks can be erased and everyone can buy the tracks that remain. It should not be to many tracks. 10 to 20 I would suggest.

Here is my list: (Bands name - Tracke name - Album name)
Helium Vola - Sabre d'amor - Für Euch, die Ihr liebt
Vicente Amigo - Ventanas Al Alma - Vivencias Imaginadas
Kodo - The Hunted - Best Of Kodo II
Patrick Vaillant - track6 - La bela naissenca
Annette Dasch - Ah crudel (Rinaldo) - Armida
Jazz Pistols - Boodubudubop - 3 on the floor
Pat Metheny - Ferry Cross the Mersey - One Quiet night
The Prodigy - Medusa's Path - Always Outnumbered, Never Outgunned
Manowar - Sons of Odin - Gods of War
Nora Jones - Little Room - Not Too Late
Quadro Nuevo - Gloomy Sunday - Tango Bitter Sweet
Orlando Consort - Ave colo vini clari - Food, Wine & Song, Music and Feasting in Renaissance Europe





Title: Re: Creating a common test library
Post by: PeterSt on December 09, 2010, 11:55:22 pm
Outstanding idea ! You may know I'm always in for awards / prizes, well you've got one from me here.

One addition maybe;
Although I don't know most of your tracks (at least not by heart by name) I guess this will be a fine list to start with. But, if you agree with me Adrian, I think when a next one comes up with a track, it should be accompanied with a description of how it ever sounded (for some part or MM:SS offset maybe), how it sounds now, and what happened in the system that caused the improvement. This may be a most nice thing to learn from for everybody. This is just an idea of mine and in the end not a necessity. On the other hand, when stories similar to what I mean are not there, everybody may be adding tracks and in the end nobody knows what they are there for. So, it acts a bit as a filter too.

Yes, I think your idea is great.
Thank you Adrian.


PS: Count me in to receive the tracks, but put me at the end of the list when you are fully booked.



Title: Re: Creating a common test library
Post by: Flecko on December 10, 2010, 12:26:50 am
Great Peter!
I will now start to load up to the internet. once the files are up there, everyone asking for them can just download them without standing in a cue. It will take some time and it wont be finished today. Then they are 2 weeks availble.

Quote
But, if you agree with me Adrian, I think when a next one comes up with a track, it should be accompanied with a description of how it ever sounded (for some part or MM:SS offset maybe), how it sounds now, and what happened in the system that caused the improvement. This may be a most nice thing to learn from for everybody. This is just an idea of mine and in the end not a necessity.

You mean, if everything is done and we all have the tracks there should be a standart how we note what we hear and in which minute an so on. Yes, sounds good.


Title: Re: Creating a common test library
Post by: Leon on December 10, 2010, 03:29:48 pm
Adrian, Peter

Indeed excelent idea, count me in for participation, I would like to add the album Touch of Yello, Peter I am sure you can point out one of the best numbers for this purpose.

Ciao
Leon


Title: Re: Creating a common test library
Post by: crisnee on December 11, 2010, 06:42:44 pm
This is a great idea, although I don't think it's meant for people like me who aren't "hardcore." Nevertheless I'd like to add my .02.

I think it would be good if more common music tracks were used, of the 13 tracks that Flecko listed, I've only heard of 2 of the artists. I have several thousand cds/records whatever, and have been in the music business to one degree or another for many years and still--granted I'm from the U.S. but again, still...

So why does it matter. Well, I'm being selfish. It's for people like me, people who don't necessarily want to join in this group but who would like to learn by osmosis (or reading and comparing what you all come up with). So if we own some of the tracks or can acquire them easily we'll be able to, and in the long run that will help everyone here that's interested in quality music/music reproduction. I also think it's good to be familiar with music that your evaluating. It's hard to truly evaluate the "musicality" of a track if one is listening to it for the first time. If one is just listening for the sounds I suppose it doesn't matter so much, although I'm not even sure of that.

What can non-hardcores learn? Well for one thing, what exactly people mean by all the descriptive terms they throw around. Now we'll have an actual reference, i.e. the bloom on that rose (the violin when it enters at 2:10) is beyond red when XX is set to .... on my system, ok I'm joking, but you get the point, I hope. Now we can compare what if anything we here at that time in that track and stash the corresponding term in our memory.

I'm also constantly amazed by how someone's audio system that was already "high end" by all accounts, can take on seven "huge," improvements in sound quality over the course of a year or two. What is the system rated now? Uber Super dupey high end? So this might also help us get a better picture of what's really going on here soundwise, how important the improvements are to us individually.

So that's my say. I hope this thing gets off the ground and you consider what I said as it will help us all over time. If you want me to clarify what I'm saying let me know, as I realize I was being a bit goofy at times (just can't be serious straight through about much).

Chris


Title: Re: Creating a common test library
Post by: pedal on December 11, 2010, 11:55:22 pm
I'm also constantly amazed by how someone's audio system that was already "high end" by all accounts, can take on seven "huge," improvements in sound quality over the course of a year or two. What is the system rated now? Uber Super dupey high end? So this might also help us get a better picture of what's really going on here soundwise, how important the improvements are to us individually.
He-he, yes, this is a quite funny observation.

When PeterSt. releases a new version it happens frequently that users give feedback like “ahhh, this is much better than the previous version! The SQ improvement is then described in big words.

From the release thread (http://www.phasure.com/index.php?board=1.0) there are about 120 updates so far. If you added all the “ahhh’s”, you might think the sum would make up a huge improvement all together. Well it doesn’t.

Not all updates were about SQ, they only addressed user interface or some other non-sound issue. Some updates actually degraded the sound. One step backwards. Other updates maybe improved the treble but after extended listening it was discovered that the bass was inferior or something. One step forward and one step backwards.

But in the end – the moral of the story – is that I, as a diehard audiophile, have to admit that “the smaller the improvement is, the bigger the describing adjective is”.

On a more serious side, keep in mind that the listening experience is a very complex cognitive process. One day I can be super sharp, really listening deep into the sound and detecting really microscopic differences. The other day I am listening to the same song, being perfectly happy with the cr*p sound coming from my car stereo. Listening is a very subjective thing. (Scientists still struggle to map the brain/aural processes).


Title: Re: Creating a common test library
Post by: crisnee on December 12, 2010, 05:15:39 am

On a more serious side, keep in mind that the listening experience is a very complex cognitive process. One day I can be super sharp, really listening deep into the sound and detecting really microscopic differences. The other day I am listening to the same song, being perfectly happy with the cr*p sound coming from my car stereo. Listening is a very subjective thing. (Scientists still struggle to map the brain/aural processes).

I agree 99.99%. I often think that a particular day's/hour's/moment's sq has more to do with my mood, the weather, my neighbor's mood, murder in Bangladesh etc. than any hardware/software. Why else could I be enthralled by a Bach violin concerto playing at Macy's through the p.a. system, when my favorite music played on my own very nice system makes me want to turn it off?

Some of my best musical memories (those that last into the next day I think Peter or someone said somewhere--how about into the next millennium) are "Big Brother and the Holding Co," Cheap Thrills, played on a Motorola? one box mono record player, and the Bach Mass in B minor played on an all Lafayette system (my first real stereo). They sounded so rich, spacious and moving. Really? Yup!

Please, no serious replies, I'm just sayin'.

For those not familiar with Lafayette. They were the first (I think) national
mid-fi--contemporary jargon unknown back then--solid state electronics and speaker producers in the U.S.

Chris


Title: Re: Creating a common test library
Post by: Flecko on December 12, 2010, 09:20:29 pm
Quote
I think it would be good if more common music tracks were used, of the 13 tracks that Flecko listed, I've only heard of 2 of the artists. I have several thousand cds/records whatever, and have been in the music business to one degree or another for many years and still--granted I'm from the U.S. but again, still...

So why does it matter. Well, I'm being selfish. It's for people like me, people who don't necessarily want to join in this group but who would like to learn by osmosis (or reading and comparing what you all come up with). So if we own some of the tracks or can acquire them easily we'll be able to, and in the long run that will help everyone here that's interested in quality music/music reproduction. I also think it's good to be familiar with music that your evaluating. It's hard to truly evaluate the "musicality" of a track if one is listening to it for the first time. If one is just listening for the sounds I suppose it doesn't matter so much, although I'm not even sure of that.

The main goal is, that all people that are interested should have the same songs on their pc to judge their pc playback. That needs records of outstanding quality. It is not my intension to force everybody to listen to my tracks. This list is just the beginning of a collection from that the best sounding records should be taken. Everyone is pleased to make suggestions. I selected the songs just by record quality. I tried to cover different styles of music but these are "just" the tracks that I have and the music I listen to. I hope very much that there will be a lot of better sounding records with different styles that I normaly not listen to. It is about the quality of the record that counts in first place. If you like to have more common tracks in the list, please tell us which songs you beleave are good and all can listen to them and decide.

Quote
This is a great idea, although I don't think it's meant for people like me who aren't "hardcore." Nevertheless I'd like to add my .02.
I think you should participate. Everybody (known) can.

I got one request by mail from a guy, that has registered two days ago, just to participate on this topic. I want everybody in the boat but if you are not known here in the forum, please tell us who you are and make yourself know first.

Greetings
Adrian


Title: Re: Creating a common test library
Post by: PeterSt on December 31, 2010, 12:38:20 am
Hey Adrian - At last some time to have a kind of response to your uploaded tracks ...

I listened to them right after you uploaded them, but "my" problem is ... I don't have remarks. Maybe one : Nora Jones needed a lower volume with her "exibiting" voice (singing louder without backing up from the mike) which may be someting not right in my system.
I am not sure whether I should say that these tracks don't contain all what can be listened for and that this is the reason I don't have remarks - or that all the recordings are just too good to ever be able to have remarks. Also it is difficult maybe to express remarks, like my example from Nora Jones. So, the only thing I'm really saying is that this will be my system (ok, from two weeks back, while ALL changed in the mean time, haha), so, do I like to say *that* ? Or would people merely say that Nora can't sing or that it's a bad recording ?

So ... I guess things must be the other way around a bit, and which is in between some lines what I suggested earlier. Example : I heard Nora Jones like this. This was on Vista. Now I play it again, and can tell that the problem has gone. Result ? Nora is a good test. "If at offset xyz she shouts, something is not good, because in my system she does not shout" (anymore !!). So, this is relative to some situation where it can be wrong.
Without such information it may just all sound nice, and what to say further.

Tomorrow (if I don't forget) I *will* play Nora again, and see whether the problem has gone with my current W7-SP1. If so, it is a nice example for each who has the problem, and thus knows what to strive for.
This is also why the examples can be about more poor quality, while outrageous quality can be squeezed out anyway. Like Stevie Wonder who sings at a pitch which can be one big shout alll over, while his "Songs in the Key of Life" is ahead of its time for recording quality. It just needs a poor volume control, that's all.

There's more examples than I can ever put up, but it needs some organization (and getting used to what we do to start with).

So far for now,
Peter


Title: Re: Creating a common test library
Post by: Flecko on January 02, 2011, 11:06:44 pm
I think the useage of the tracks depends on how you are testing and judging what you hear. My methode to judge is to listen to records with most handmade music. I listen to the instruments on the record and try to compare them to the sound I know (from memory) from the real instrument. If there is something strange in the sound, if it sounds distorted there is something wrong. I chose the setting which brings me the closest to the natural sound of the instruments. That is the purpose to use high quality records. If I tune my system to a recording of a acoustic guitar that contains distortions (not heavy metal records ;)), I will tune in the wrong direction because I will try to mask the distortion of the record.
It would be the most easy, if we first share the music. I think the best tracks will sort out by itself. Everybody judges his system in a different way. So if "you" do it in a different way, just say to what "we" should listen to. We will see which tracks will be used. Important is, that we all have them to compare.