I’m currently sequencing Schumann’s Cello Concerto, but as always, I’m struggling with adjusting the delay in the attack of the string samples. Even when I input the notes precisely on the grid according to their note values, the actual attacks end up delayed and out of sync. This issue is particularly pronounced with legato articulations. To some extent, this kind of adjustment is unavoidable when working with sampled instruments. If the delay were consistent across all notes, I could simply use the DAW’s track delay function. However, the amount of delay varies significantly depending on the articulation. It seems to be an inevitable part of working with sampled instruments—connecting individual sampled notes in a DAW to shape a performance. For short pieces, I can manage by making adjustments one by one, but when working on an entire concerto, manually tweaking every note becomes an extremely time-consuming task. And since I have five string tracks, from violin to double bass, the workload multiplies by five. Woodwinds and brass instruments have similar issues. Fortunately, they generally have fewer passages compared to strings, so it’s somewhat tolerable. Various techniques have been suggested to address this problem, but none seem to work quite perfectly.
On the other hand, physically modeled instruments, which have been improving significantly in recent years, don’t have this issue—they align perfectly with the grid. That’s the advantage of being entirely computer-generated. Woodwinds and brass sound fairly good, but when it comes to strings, they still fall short of the rich tonal nuances of real string instruments. The sound tends to be somewhat monotonous and synthetic. I hope they can push the technology a bit further.
However, a Japanese-made sampled string library I recently found offers a solution to this difficult problem. It’s called Tokyo Scoring Strings, and the key feature addressing this issue is Look Ahead (see the oval-marked section in the image below). For example, in legato performances, this function anticipates the phrase and legato transitions, ensuring that the next note-on aligns precisely with the intended pitch—without requiring manual note position adjustments. And it does this in a natural way. As a result, quantized MIDI data is played back in perfect sync with the grid. To achieve this, the system applies a one-second latency to allow for pre-analysis and adjustment. I found this to be quite an ingenious idea—one second should be plenty of time for detailed processing.
I am able to type quite easily with this sound library. The sound quality is probably to each person's taste, but it is fine for me.