Jyri T. wrote:The navigation between sections must be a one-clicker. There are two bars visible, why don't you let the user to go from one section to any other instantly. Paavo said earlier that is it so difficult to click twice and I say yes, because a) the designated button is microscopic and, more substantially, b) this way it is really difficult to compare two sections (since you have to jump from a section to whole song and then back to the other section). And for crying out loud, what's the point in deliberately designing your program to be intricate?
I would certainly not be so arrogant to say that there is no room for improvement in the segment navigation though I don't think it is as difficult as you have described.
Lets say you've clicked on the first segment and it is zoomed into it but you now want to navigate to a different one. I presume you know which one you want to navigate to so the way I do it is to click and drag the "visible zone" on the upper bar so that it is in the segment you want to move to, then release the mouse and click on the bottom bar. So it's one drag and one mouse click operation. Certainly not a lot of effort in my book and I don't see it as intricate.
The main problem as I see it with the time-line control is not a lack of functionality but a lack of knowledge of the functionality. At some point I plan to add hover over animation to the bits you can click on and drag to make it more self evident but I don't think there is a great deal missing from it.
On the issue of comparing sections, that too is not hard. Again, if you've selected one segment but now wish to compare it with another drag the "visble zone" to the segment you want to reference, then click and drag the "R" in the top right hand corner down to the bottom bar and release the mouse button and there you have it. Again, it's just two drag operations without any need to de-select the current segment.
Jyri T. wrote:I have been able to produce some hickups:
- HB3 crashed on me after I tried to open an HB2 filter.
- The loudness matching seems to be at least occasionally up to a few decibels off (or I unknowingly messed up things).
- The filters acted weirdly a couple of times. I opened a filter and as soon as I engaged it, the limiter indicator jumped to the roof and the song would play like it was compressed to instant death and then some. Reopening the file the second time stopped the weirdness.
- The playback has been acting funny a couple of times (e.g., it sounds thinner that should, after stop and re-play sounds good again).
- Automatic spectrum resizing after an adjustment is a major PITA! Just when you locate the problem part of the spectrum, the program decides to change the scale. I want an option to turn it off!
- The navigation bars act funnily if you click on the upper one while working on a section.
HB3 does not understand HB2 filters. I did think of adding some support for it but decided against it because the differences between the two are huge and I didn't see much benefit in it. If you try to open them it won't work, but I shouldn't crash either so I should fix that.
Loudness matching is more than just matching levels. That little something extra is not something easily codified (not by me at least) so automated loudness matching will never be perfect. I noticed similar issues myself when re-coding this for HB3 but I spent a great deal of time testing it and found nothing wrong with the code. I recall a case of matching levels between a modestly compressed track with a completely squashed track and finding the squashed track sounded 2dB quieter even though the average levels were the same! That probably stems from the squashed dynamics, yet another reason to avoid it.
I've made a fix to do with file locking (I don't know if it is in your build) that may explain some of the errant behaviour though I have a feeling I still have some cryptic ones left to solve.
As for the remainder, I'm on the verge of burn out having been working on the code for how many years I don't know, so I am very reticent to do anything substantial to the current interface until after a release and a holiday. You'll have to wait until the dust has settled on HB3 before much UI optimisation takes place.
Release is delayed in any case, because in my porting to Mac I have uncovered some file format related design issues that needed fixing and the experience has left me set on not releasing until after the Mac port is functional to ensure that I don't release a malformed format and then have to deal with the design fault later. I also lost quite a bit of time settling on a new build environment for the Windows platform (I've now switched to Visual Studio 10) and then having to deal with STL and ANSI C library issues with Microsoft's latest implementations. I've literally lost a whole month dealing with this rubbish but I'm finally going forward again.
cheers,
Paavo.