One Synth Challenge #105 - Thorn by Dmitry Sches (J. Ruegg Wins!)

[Frostline]

I've cross-checked the data listed under my name with my actual vote...
The scores shown on the chart is definitely not what i voted for.
I even took screenshots, so i know exactly what i voted.

The data under your name in the top row is the score the people gave you.
The data next to your name in the left hand column should match what you voted.
If it doesn't match correctly this could be a big issue with the system.

[sqigls]

The data in the left hand column is not what i voted.

[Frostline]

. didn't want to open a needless can of pointless internet debate!

This is maybe the 3rd time this conversation has happened in my brief time here at OSC and I'm confident it will not be the last.

Personally I start at the bottom of the scale. what does it take to get a 1?
Well technically holding a single note for 2 minutes on the default patch would qualify as a 1.
So to me if a track does more than that it at minimum has to be a 2 just to separate it from total lack of "anything" in the 1 rank. Which is why I have yet to give a 1 (other than to my own tracks of course and even when an artist requested a 1 from everyone to prove some point I didn't not grant it because it is my voting system). So yeah I look at 1's being pretty dog poopy too.

But look back at previous months, it is amazing that there are always tracks in the top 5 with 1's and 2's. I have no idea what they are not hearing in the work.

My last entry I got what I considered 3 low scores(relative to what the other participants granted my track) yet I did "better" overall in ranking than those 3 people. So take satisfaction that those who give you low scores really might have no clue since they didn't manage to crack the top spots themselves.

[Frostline]

The data in the left hand column is not what i voted.

That is very concerning.
Did you have any issues during voting? Couldn't log-in? Did you refresh the page after completing voting to make sure all votes stayed? (I've had a couple of times after a refresh a few tracks suddenly went back to being unvoted).
In the past I tracked what I voted with the data sheet and it always matched so I stopped doing that.
Next voting I will definitely go back to careful scrutiny.
Hopefully it was some easy to correct error in the process. I don't know how the voting page data gets entered into the Google spreadsheet whether it is an automated process or there is a copy/paste of the data.

[generaldiomedes]

My votes are 100% correct, make sure you are looking at the row with your name, not the column ..

[sqigls]

just logged in again, my vote is still there as i left it, which is different to the results on the chart.

[] Peter:H []

just logged in again, my vote is still there as i left it, which is different to the results on the chart.

Probably there's some glitches in the voting software? I remember that I came back once to adjust some votes and found that some tracks moved again to the "not voted yet" column. Some times later other users reported the same issue...just guessing here.

Anyways voting about art is almost impossible...I'm in software architect and I did a lot of Architecture Reviews where I looked on quality aspects of other peoples software...doing such I used the Goal Question Metric (GQM) Method. You have a Goal you want to achive with each Review and certain Questions will guide you to the answers which are backed by a set of metrics you collect. Could you apply the same tomusic? I say no, read why...

Most of the times I found a completely subjective point of view in Goal and Question even in Software Architecture. And Goal and Question even changed over time. One sponsore is today probably interested in performance of the software (what ever "performance" means) the other sponsor is tomorrow interested in maintainability (whatever that is). So goal and questions is allways to some extend individualy based on a point of view of an individual. It was hard to consult my victims to find the questions they actually really have...
The last level "Metrics" is usualy more interesting, as you can spend some time to find objective, reliable, reproducable, fact based metrics, even potentialy based on a data colletion of an gauged sensor...

But not here. Musik is art. What metrics would you collect? Frequency density? I bet that Arnold Schönberg would never ever win an OSC, same as John Cage for http://www.nytimes.com/2006/05/05/arts/ ... 0&emc=eta1. Both are respected artist...

I for instance voted this time without any goal but a single question "what track sounds interesting and not like to much commercial of the shelf?" (I say "I" like in this is totaly subjective point of view, please don't discuss my voting with me...If you do, you should reread this post)

However - All this "investigative" approach to the voting of other people is an completely futile attempt to project someones GQM system onto another person...it won't work up until the same metrics are a.) given for all and b.) meet the quality standards as mentioned above...

That said...just have fun...don't be botherd by a place in top five...some people not even get close to these ranks...positive aspect...if you're on top this time...next round you can not be better only the same...most probably you'e not on top again...what a shame

[functionform]

Thanks Frostline! iirc that comes out to almost exactly what I got last osc! I'm going to start charting them. For me, its not about whether I win or not, but whether I grow with respect fellow participants likability. The comments that people leave are huge for feedback purposes, and everyone here is so helpful. The joy is totally in participation and growth.

Wrt to scoring, I came to a similar sort of system as you mmghost, but I didn't think to add in the "would you know its an osc" critieria, I do like that a lot.

[zarf]

Voting Conversation Alert - Skip if Not Interested

This is how i go about voting.
5 - Track stands on its own, would never know it was part of a one-synth contest, sounds professional, could be released on a label
4 - About the same as a 5, but perhaps lacks a professional mix-down, or perhaps some of the sounds give it away as a once-synth track
3 - Middle of the pack, I often find that 3's are solid but often sound as made with one-synth, or perhaps lack a bit of tension and release in the composition, no "ear worms" or other elements that have me saying "how'd he/she do that?"
2 - Track has some nice ideas, but either has elements that hurt my ears, or perhaps was not well-thought out from a composition standpoint.
1 - Someone recording them self messing with presets, no thought at all given to the music or how to impact someone emotionally

That's interesting. While my votes might often end up looking pretty similar to that, my strategy is actually pretty different...
You have described one version of an 'absolute' system in which you evaluate the tracks against some criteria (nicely described by you here - helpful )

For me - of course I still assess tracks for how 'good' they are (sound design, arrangement, production, feeling/interest/magic) but the actual score is all relative to the other tracks that month. The ones I think are best get 5s and it works down from there. Sometimes 3 'buckets' is enough and I feel like it's good enough to just use 4s and 3s... other months, when there is a really broad range, I might need to use 2s and even 1s sometimes - though I try to avoid that just because people project that that means i think 'It's crap' even though that is usually not the case for me.

Bottom line - a 1 or a 2 doesn't mean I think it is 'bad' or 'dog crap' for me, it simply means that
The 1s were not as good as the 2s which were not as good as the 3s which were not as good as the 4s which were not as good as the 5s - IMO, of course

So in a good month with loads of great entries, even the tracks I occasionally put in the 1 or 2 bucket, I still can think they are pretty good! Just weak relative to the competition.

Just how I do it

[liqih]

...

Thanks for sharing your thoughts.
What you wrote is very close to what I think in terms of metrics in liberal arts, technical and aesthetic qualities are not separable, hence any review is highly subjective.

[generaldiomedes]

Here's how I do it.

If I don't like the genre, -1.
If I really don't like the genre today, -2.
If you've been trolling the forums lately, -1.
If you voted me 3 points or lower last contest, -1.
If you left me a passive aggressive comment last contest, -1.
If you left me a nice comment last contest, +1.
If you scored me 5 points last contest, +1.
If I just think you need to be taken down a few notches, -1.

I am kidding of course! I am actually very close to mmGhosts method of scoring, I just don't have a 'label sound' bar for a five pointer.

[doctorbob]

I'm very much like Zarf with my scoring! I do put some emphasis on composition too (but then, I would, wouldn't I?). I try to be objective, but inevitably, some subjective criteria will creep in. That is, in general, I have to "like" the higher rated tracks as well.

I can't pretend to "understand" all the genres we get here - but (some will be surprised), I have done my fair share of electronic bleeps and bloops styled tracks, but way back in the late 60's and early 70's (and then got drawn away by the prog rock stuff for a while). Still play my David Vorhaus White Noise albums (vinyl, of course!) now and then, along with Curved Air and Terry Riley.

dB

[z.prime]

You know... The best way to score things is just to look at peoples names and then score them accordingly

[Hanz Meyzer]

Well, I commonly hate dubstep. But! Not because of the genre - not at all. Even for me, there are some fine, really impressive tunes (usually the older the better). It seems to me that the current standards in that genre are totally f**ked up, like way too loud, way too harsh, no dynamics at all - there are some exceptions surely. Esp. the drums sound too wide-spectrum eqed to me. If there was a dubstep song with smooth transients, no sirens, no rave-like snare rolls, no obvious hymn, I surely would vote it 5 stars, I like modern music in general. The same happened to psytrance genre, it is a dead sounding genre now - So I can't help it: Specific today's genre standards come with specific dynamics, eqing and standard sounds, and that's what I do not like about it. (And why dubstep today always needs that always-same-totally-chavvy-sounding snare sound?? WHY! Can't you use another sound, so even a Hanz Meyzer could enjoy the song? It is ugly!)

Of course this all is totally subjective, and also I seem to be the only one thinking like this. Btw. it seems to me that the harshness of the songs constantly drops with every round in KVR OSC, which is to be a good thing to me.

What I actually wanted to write: If it sounds harsh to me, I can't vote it 5 stars, even if the composition, layering, editing, modulation is outstanding work. So I usually vote the overall sound and mood - which of course is highly subjective. Is that wrong or unfair?

[doctorbob]

You know... The best way to score things is just to look at peoples names and then score them accordingly

... and probably some people do vote this way - especially since we all hide behind "KVR usernames"
Seriously though, I do listen to the tracks and try to ignore who made it ... but, some folk have "signature" sounds

dB