Mixed In Key 7

I have tested it the last 2 days to quite some extend and I have to say I am really impressed with the automatic cue points.
Now of course they are not for every song, so if you have a special routine planned with a song or a scratch phrase, this is where manual cue points shine. But for the “usual” songs where you need things like 1st beat, 1st breakdown, 1st chorus,…. they are golden and hit it quite good most of the time (I would say 90%+).

Also I think the key detection and energy level detection made another step forward. It is hard to say as they were already very good and the last few percent are hard to detect. So this one is more like a feeling.

Hmm I tried mostly house and trance tracks. Did you have problems with other genres? It looks to me as the success rates gets less good with hip hop for example (just tested a few tracks).

What I am finding is most of the songs will be off by 1/8th of a beat before or after the 1st downbeat. Sometimes the cue points are off by a 1/2 beat or even a 1/2 measure after the first downbeat. I even have some songs where the first cue point is way before the music even starts. All these issue are on songs that have correct beatgrids.

I did notice when Mixed In Key gets it right it does start on the beginning of a phrase.For example it will set cue points on the 1st down beat of measure 1,9,17,25… and so on.

One of the problems I see with this feature is if the song beatgrid is not analysed correctly it will drop the first cue point on the start of the music wherever that beat may start.

I know this is not going to be perfect and I don’t expect it to be. If anything I will use it more as suggestions on where to place my cue points and edit as I go along.

I have to admit I was quite excited when I first saw the auto cue-point functionality as I tend to place 75% of cue points in largely the same places in every track.

My hope was that the auto function would put the common ones’ in and then I could customise as needed. From the above posts it would seem some people have been reasonable happy but unfortunately I haven’t had the same experience. For me, there are two things stopping this from being an amazing tool:

1. The detection algorithm seems to consistently miss the start of a phrase by around 1/2 a beat. It did this every time, on pretty standardly-structured Deep House (Serato had no trouble getting the grids in the right place)

2. There doesn’t seem to be a pattern to where it chooses to put the cue-points. I would have liked to have at least the first few placed at common basic mix-in points (e.g. 9, 17, 25, 33). What I don’t want is cues placed slap-bang in the middle of the breakdown – as happened on a few occasions. It’s all subjective of course, which is why it probably needs some customisation features. Perhaps the ability to preset where each auto cue is placed based on phrases? And maybe the ability to select the number of cue points set, so that it doesn’t fill the whole bank up with cues you don’t need?

Of course this is just v1 of the tool and I’m sure it will improve over time, but it’s a shame it’s not particularly usable at this stage (to me anyway).

I don’t mean to be too harsh, but I have had to resort to laughing off the complete failure MIK has had with my music. I do play a lot of hip hop, but also plenty of electronic stuff as well. So far, I’ve had 0% success rate. As I said, it’s pretty much been amusing how off the cue points have been – totally off the 1st beat of the bar, placed bizarrely in the middle of a verse or breakdown, completely ignoring song structure seemingly… I look forward to the software developing, but at this point it seems there is a long way to go.