josq wrote:Dear Sonar11, I have said that these statistics have a lot of limitations, and I'm aware of the one you mention, and I think that's true for everyone here.
Nevertheless, I should have added "on contrebombarde" to the title of this thread.
No one has to attach any conclusion to this list, but one interesting conclusing could be how much top contributors skew the representation of sample sets. It may make them interested in using different sample sets when possible - there are several good quality sets which really deserve more attention.
On the other hand, among the top uploaders there are a few very good organists. So if you consider to buy Caen, Rotterdam, or Utrecht, it may be ensuring that these sets are so popular in the context of Contrebombarde.
I don't think you get it; adding "on contrebombarde" to the title changes nothing. I don't have a problem (necessarily) using contrebombarde data, but you're using it incorrectly. Here is an example; Let's say tomorrow I go out and purchase a new ford focus (car), and tomorrow you go out and purchase a honda civic. Because I live in the country in North America, I use my focus every day, back and forth to work, to visit people etc etc. Because (let's say) you live in Europe with better public transportation, you only need to use your civic once a week. Now in April some car fans get together and try to draw conclusions; "the focus must be a more popular car than the civic because it is used more often in the month by me". Does that seem anywhere close to accurate? No, it isn't, because we both use our cars.
A person using a sample on a monthly basis should be a binary count; "yes", or "no".
Finally, you still seem to be under the impression that you can use popularity to define quality. You can't ever use popularity to infer quality, for anything, not just HW. That is a common misconception. Even if your statistics were correct, and even if those 25 samples were ranked correctly, none of that in any way means that "Rotterdam" must be higher quality than "Tholen". As I said in my previous post, a set can become popular for many reasons that have nothing to do with quality. Here is a hint (and possible reason): What do those 3 samples you listed (Caen, Rotterdam, Utrecht) all have in common? Well yes they are all Sonus Paradisi (and I agree that his sets are high quality, no problem there), but they are all famous organs in real life. Wouldn't that have a huge impact on "desire" to have the HW version of that organ? Compared to say, some less famous organ in Tholen, Smecno, or Zoblitz?
I will stop arguing, and I apologize to you for making an issue out of this (I'm not trying to attack you personally), but I wanted to post this because I am absolutely against publicly posting "these sets are higher quality than those sets" types of statements. You may form your own opinion of course, but I don't like public comparisons of X vs Y when it comes to sample set producers. HW is too small, and we need the efforts of all producers to keep this incredible software moving foward.
Edit: after reading your last post again, I also wanted to mention one other issue I have with your post; just because an organist is talented, it doesn't mean he automatically has the ability to pick out high quality sets. I know of several people who can barely play organ, but fix (or build) organs for a living and can easily pick out flaws in a sample. It also works the other way, I also know of organists far better than myself, but who don't necessarily know anything about sound, recording etc. Hopefully now you understand why I have so much trouble with what you've posted here.