
Sometimes when people find differences I get to hear about it. In such cases I will always investigate, but I can't do so if nobody explains what the difficulty is.

The first time someone told me they found differences between Yoshi and Zyn, they were actually thoughtful enough to supply recordings. Imagine my puzzlement when I was unable to reproduce either of these! By degrees I worked through possible explanations asking about the other's setup and eventually got round to asking what he had in 'Settings'. At this point he realised they were quite different (and different to mine too) - problem solved.

Later I had another complaint, but this was much harder to pin down. The person was working from the virtual keyboard so I did the same. Settings seemed identical yet the difference was very noticeable, really quite obvious. Eventually I started to deconstruct the instrument patch bit-by-bit and the difference suddenly disappeared when I disabled the Addsynth voice 1 filter envelope. Inspection revealed that this had very extreme settings. It still took me a long time to hit on the answer. After going through the code in great detail I could find no significant differences. The breakthrough came when I realised that the filter behaviour was critically dependent on the starting velocity, and *only* in the virtual keyboard were the *default* values different. The next commit of Yoshi had this nailed. However, had the settings been checked, exactly the same sound could have been achieved.

Some years ago someone sent in a patch with spelling corrections. Unfortunately this changed the name of an XML parameter. It was quickly found and corrected. Ironically the only patch I found that was affected by this was one of mine!

More recently we discovered that the jack and internal buffers had somehow become fused together. This would make a difference to some sounds if you wanted a very large internal buffer but a small jack one, although it's a situation I never came across myself. However, this in now corrected.

The very latest (quite obscure) sound difference turned out to be a regression in Zyn. This has been fixed.


Whenever in doubt my reference is to check sounds against Zyn 2.2.1. This is getting harder as I can no longer compile it on modern distros, and the executable will save but not load instrument files. Also, some features like 'unison' don't exist.

All-in-all, my experience is that provided you set up *exactly* the same configuration in all respects an instrument created with old Zyn will sound virtually the same on both new Zyn and the current version of Yoshi.

There is an extra twist to this though. I'm the sort of of person that doesn't throw things away until I can no longer shut the cupboard doors. I have quite a few computers of varying vintage and have noticed that there are small, but noticeable differences between these with both Yosh and Zyn. At first I thought it was due to different distros, but by degrees I've brought them to pretty much the same standard. Therefore my conclusion is that it is down to architecture variations.


Finally, wearing my musician's hat, to me it is more important to get consistently the same sounds on my setup than for mine to sound the same as anyone else.