Synthesizer V Studio 1.8.0 更新 / Update / アップデート

检查更新 / Check for Update / アップデートを確認:

——————————
#lang:zh

SynthV Studio 发布 1.8.0 正式版,相较于公测版本除了大量 BUG 修复,还改进了 DPM 音素力度的调整效果,同时插件版本也支持用 MIDI 键盘录制音符。

所有的 AI 歌声数据库都正式迎来 DPM 模型的更新,极大改善了自动音高和音色的表现(Mai 本身就基于 DPM 模型所以不需要更新)。

更新日志:


此外夏语遥为了针对 DPM 模型进行改进,宣布延期推出。预计测试版在 12 月中旬发布,而正式版将在明年 1 月初推出。

——————————
#lang:ja

更新履歴:


——————————
#lang:en

Update Log:


——————————
Previous version

2 Likes

Hi, I am still on 1.6.1 because of distorted vocals in English voices when going to 1.7.0 and 1.7.1. My old songs from 1.6.1 where UNusable after those updates. I learned from my initial update to make a copy of the folder containing the voices and the settings, so I was able to retrieve my work. Now I wonder if this problem has been finally repaired. Does 1.8.0 cause anybody problems with songs created in 1.6.1?( Distortion , wild vibratos that were not there in 1.6.1.) I am very hesitant to upgrade although I have already bought Natalie ( Not knowing then that it Needed 1.7.1) and could really benefit. Any feedback most welcome! Thanks!

You can try to update the latest version first. And in the latest version, when the voicebank is updated, the old version of the voicebanks is also retained, so there is no big concern.

Thanks. So if I understand things correctly the VOICEBANKS caused the distortion and vibrato problems? And NOT the new HD voice code from Synthesizer V itself? That would be a relief.

In the latest version, the model has been updated and publicly tested for some time, so I think the previous problem has been solved.

1 Like

There are very small differences between using Solaria 103 in version 1.6.1 and 1.8.0 of SynthV Studio, but they’re very minute, not nearly as big a difference as some people saw with 1.7.x and HDVM. This is likely thanks to Dreamtonics adopting a beta release process and learning from past mistakes.

The changes in sound are more pronounced when you move the actual voice database to Solaria 108, but as was mentioned you can now install both versions of the voice database simultaneously and toggle between them.

The general reactions I’ve seen on Twitter for Solaria specifically indicate that 103 still has the most “Power” but 108 is an improvement for all/most other styles.

Of course this may be different for each voice, Solaria is just one of the more prominent examples since she’s so well-regarded and people are therefore more likely to notice when she sounds different.

All that said, some users have experienced odd vibrato patterns with version 1.8.0. I haven’t personally seen this behavior yet, but there do still seem to be some quirks.

1 Like

Since the vibrato anomaly was one of the reasons I reverted back to 1.6.1 I am really afraid to upgrade again when I read this. The amount of vibrato might be applicable for Japanese and Chinese vocals but is absolutely a no-go for English vocals. I think I will have to wait for Natalie just a bit longer until this is cleared up. Thanks for all your input !!

My impression was that the 1.7.0 issues were around inconsistent dynamics, not vibrato. I’m fairly certain the bug mentioned above (which very few users are encountering) is not related to the widespread dynamics problems people had with the past release.

1 Like

Thanks. I did report some wild vibratos at the time but only in English voices. I distinctly remember starting to tediously “undo” all those when I realised I was making matters worse and reverted back to 1.6.1 after consulting with the lovely people here, Since I use solaria and Eleanor AI’s in 7 songs that are to be mastered soon as a unit I absolutely need consistency now and will probably only switch after all that is done. I could render ( bounce) them now but I have noticed that the flexibility up until the very last minute in changing lyrics and intonations its what brought me here in the first place. It is a fantastic product and I LOVE using it. Thanks for chiming in!

1 Like

Hello and thanks again for all the feedback. I prepared a return path to 1.6.1 just in case and have upgraded to 1.8.0. Almost all vocals sound as they are supposed to with the old Voice Databases installed. That is great. A brief question : If I update the voice databases will the old ones remain in the same folder ? Or do they get a separate directory. I am asking because of my backups.

If you install additional versions of an already-installed voice database they will be located in the same folder that is already being used for the existing version’s installation.

The version used for each track/group can then be selected from the Voice panel.

image

3 Likes