• We're currently having issues with our e-mail system. Anything requiring e-mail validation (2FA, forgotten passwords, etc.) requires to be changed manually at the moment. Please reach out via the Contact Us form if you require any assistance.
I
Reaction score
609

Joined

Profile posts Latest activity Postings Usernames About

  • So... Some sort of distant cousin of mine (whom I've never met) was just cast as a super villain in the MCU. Good for him!
    WyndReed
    WyndReed
    Wicked awesome!
    I
    inactive
    In case anyone is wondering, he'll be playing Adam Warlock in Guardians of the Galaxy Vol. 3. I'm sure you can figure out the rest of the details from that.

    The funny thing is, Will Poulter is from my dad's side of the family, and my dad has been keen to watch all of his films. However, my dad's not exactly an MCU fan, but now he will have to endure an MCU movie whether he likes it or not!
    MillyAqualine
    MillyAqualine
    Holy milk !!!
    I just finished work on an Eleanor Forte song, but it won't be "officially" released until I can replace the vocals with the AI version of Eleanor Forte. I guess that means October at the earliest. (Composition began in late January.)

    At any rate, here it is:
    Dream...

    The fast-paced, lower-pitched vocals are difficult to understand. Hopefully this won't be an issue with AI Eleanor. (Absolutely no tuning was done for this current version.)

    It's a Minecraft song, and the words and music were written side-by-side, with each pushing and pulling at one another. I'm quite happy with how it turned out. My favourite verse is this:

    Cull some sheep
    And craft yourself a bed from murder
    Cull some sheep
    And go to sleep...
    So I created a forum thread on guitar cabinet impulse responses. Why?

    About 18 months ago I wrote a guide on how to use free Ignite Amps software, but since then my knowledge of amp sims has improved. (Not to mention that better free alternatives have also emerged.) I began pondering if it would be worthwhile to clear up any explanations or fix up any errors in that guide. My initial instinct was to not bother, after all, nobody has ever responded to that thread. Perhaps not surprising when considering that amps sims aren’t exactly this website’s raison d'être.

    However...

    Much to my surprise that guide has over three-thousand views! Who’s reading it? Probably not the VocaVerse regulars, so it must be casual visitors.

    So I decided to write an entirely new thread on a similar subject just to see what happens. The only thing for me to do now is wait 18 months and then check the view count.
    mobius017
    mobius017
    I have a suspicion that there are lots of people who come here who aren't members, browsing the threads for news or the resources for info.
    Another excellent Dan Worrall video, but this time it's not about audio plugins and mixing. And I'm sure that many on this forum will be able to relate directly.
    I'm beginning to suspect the "harshness" in Saki AI, which is still audible in version 104, is aliasing distortion.
    I
    inactive
    It's not so much that Saki is difficult to mix, but that as you go higher in pitch, the more harsh she sounds. In fact, just the observation that it gets worse with higher notes is further evidence that this may be some sort of aliasing distortion. And this is not a problem limited to Saki AI, as I first noticed this in the original SynthV using both Eleanor Forte and Yamine Renri.

    But to answer your question, yes, the audio is being distorted (or whatever) at SynthV's end.

    Additionally, I've since learned that the "mirroring" effect I saw in some of my prior tests is called "folding" and that it can also be evidence of aliasing. At the time, when I first saw that folding, I wondered if it were related to foldback distortion (another name for audio aliasing), and now I know that it is related. It seems to be an unfolded type of foldback distortion. (Don't ask me for details because I don't have 'em.)

    Regardless of whether or not I'm hearing aliasing, there is something wrong with SynthV's higher notes. Back when I goofed around with Hatsune Miku V2 and V4, you could easily go an octave above the recommended range without any harshness. Sure, it would be squeaky like a lung-full of helium, but it wasn't brittle. I can't do the same with SynthV without blood dripping out of my ears.

    It's kind of a shame because I think SynthV has some of the best underlying technology out there. And if there is an aliasing issue, Mr. Hua has obviously put it very far down on his list of priorities.

    On the other hand, if you don't go above D5 you're probably okay.
    Ronny777
    Ronny777
    The way you explain it makes me think that this is what I am encountering when I tune with Miku english. I found that by using singing skills or Attack and Release effects on every note it give the voice a "sand in mouth" type of sound when it goes too high of too low. So, maybe it's the AI technology itself. Like the Attack Release from V5 the AI from synthV might be adding too much to the voice. I really don't know how SynthV works but if she was fine before the AI ver. of her voice, this the most logical thing I can think of. With V5 you can tone down the amount of effects and singing skill but can you do that with SynthV?
    I
    inactive
    Harshness issues have been with SynthV since before SynthV Studio was ever a thing. It is unrelated to the AI features.
    I recently started working on a book review for school. The book was only available from my university's library as an eBook, which is unfortunate because I prefer reading physical copies. Nevertheless, I read it last week, and yesterday I typed in all the points I wanted to discuss in my review.

    This morning I began researching the author to learn more about their qualifications. However, while doing so, I discovered why the book was only available electronically: the publisher had recalled and destroyed all unsold physical copies. The reason for this was because large swaths had been plagiarized from other sources.

    Ugh. I may have to read another book for my review.

    Moral of the story: Remember to check your sources!

    And as is usual in plagiarism cases, this guy had a history. I wonder if his PhD was revoked?
    mobius017
    mobius017
    I think it'd be worth bringing the issue up to your professor and seeing what he/she says. Just because the book you're reviewing (Which I assume is like a critique or book report?) was produced using improper means doesn't mean you can't write a fully proper review of that book, at least to my mind. You could even bring up the plagiarism in your review.

    Unless a "book review" means something different than I'm thinking. Worth checking to be sure, I think, just to avoid the loss of the time/effort you've spent so far.
    I
    inactive
    I've already discussed this with the professor and he just let me know that I can continue on the current trajectory. However, once I'm finished with the review, we have to alert the library and get the eBook removed.
    Blue Of Mind
    Blue Of Mind
    I had to do a book review during my first year of uni, and I didn't particularly enjoy it because I'd never done a review before. (Yes, not even for school). But I find the idea of reviewing a plagiarised book really interesting from an analytical perspective. You could really go into the legal and ethical issues surrounding plagiarism.
    I have three songs underway (and a fourth planned) in anticipation of Eleanor Forte AI being an actual thing by the time my current school term is finished.
    Alright, I'm going to start uploading my SynthV sample rate "investigation" in piecemeal. It could take a couple of hours before all parts are up.
    I'm doing some "scientific" investigations into SynthV Studio Saki AI and sample rates. But before I finish, I must do some school work. Nevertheless, expect some informative results later today!
    Does anyone have experience with mini MIDI keyboards? I'm thinking of purchasing one with full-sized keys, but with only a couple of octaves so that it can sit comfortably in front of my monitors.
    I
    inactive
    Yeah, I'm just looking for notation input. I already have an 88-key piano for composition. But I'm wondering if one with twizzly knobs might also work with EQ plugins and the like so I can also use it when mixing.
    Älfa Dröttning
    Älfa Dröttning
    I got an Arturia Keylab MKII and love it. Mine is 61 keys, but they also have a 49 key version. It has multiple banks of knobs and faders which is nice for adjusting plugins, and it works with a bunch of different DAWs. It also has different modes for DAW control vs instrument control and it comes with Analog Lab which has a bunch of different instrument presets. The price point is on the higher end for midi keyboards, but Arturia also has a Keylab Essential version which is cheaper and doesn’t have as many features. It still has the knobs, faders, and DAW control, but I’ve read that the key feel isn’t as good. I did a ton of digging on MIDI keyboards last year and I found Sanjay C’s channel to be really helpful: https://youtube.com/c/SanjayC

    Imo you’d probably be better off with at least a 49 key keyboard than a 25 key one since you’re looking for notation input. Most keyboards have octave up/down buttons, but it’s easier to input notes if the octave change isn’t an issue.
    I
    inactive
    Whilst looking at the Arturia website I realized that space was an issue—my desk isn't exactly gargantuan. After some measurements I settled on the Arturia Keystep, and it does indeed fit snuggly atop my workspace. Plus, it has a nice solid metal base!
    I just discovered a cure for SynthV Studio’s harshness. (Or at least a cure for Saki AI, but I suspect that this works for all SynthV Studio voicebanks, including those running in the original SynthV, which also exhibited harshness.)

    Normally I use Z-Noise to try and reduce some of SynthV’s harshness, but today I got thinking. Does Z-Noise work better at higher sample rates? So instead of exporting Saki AI at 44.1 kHz, I decided to try 88.2 kHz—except there was no such option. All I could choose from was 44.1, 48, and 96 kHz. So I gave 96 kHz a spin.

    But after doing so, and even before I engaged Z-Noise, I noticed how much cleaner Saki AI had become. The harshness had disappeared without the need for extra post-processing!

    Owing to the lack of an 88.2 kHz option, I then wondered if SynthV preferred multiples of 48 kHz, so I tried exporting at that sample rate. But when doing so, I noticed that SynthV did not re-render the audio before saving. In fact, once the vocal track had been internally rendered, it could be saved as 44.1, 48, and 96 kHz in rapid succession without any extra CPU grinding. This led me to wonder if SynthV is always running internally at 96 kHz, and if SynthV’s harshness is caused by poor sample-rate conversion.

    A quick listen to the 48 kHz export option seemed to confirm this, as it sounded as nearly harsh as the 44.1 kHz export. So I tried down-sampling the 96 kHz export to 44.1 kHz using Audacity and compared it to the SynthV native 44.1 kHz export. The difference was not subtle, with the 44.1 kHz Audacity output sounding essentially identical to the 96 kHz source.

    If you are a SynthV user, I recommend exporting at 96 kHz and then down-converting to 44.1 kHz in either Audacity, your DAW, or whatever. Just avoid exporting natively at 44.1 or 48 kHz and you’ll have much less harsh sounding vocals. The difference is very much audible.

    And yes, sample-rate conversion (SRC) does vary from software to software, as evidenced by the data presented at this website:
    peaches2217
    peaches2217
    Would you be able to specify how to down-convert, or at least point to resources that help with it? I tried down-converting according to a few guides on the internet and heard no difference at all, honestly.
    I
    inactive
    Spine-tingling step-by-step guide:
    • Export your SynthV vocal track as 24-bit 96 kHz audio.
    • Open the resultant .wav in Audacity.
    • Change your Audacity settings:
      • Go to Edit > Preferences > Quality
      • Change the "High-quality Conversion" settings to:
        • Sample Rate Converter: Best Quality (Slowest)
        • Dither: None
      • Click "OK"
    • Go to Tracks > Resample
      • Select "44100" from the drop-down menu.
      • Click "OK"
    • Once the processing is complete, go to the bottom left of Audacity and change "Project Rate (Hz)" to "44100" via the drop-down menu.
    • Go to File > Export > Save as WAV
      • Make sure the encoding is set to 24-bit before saving. (Unless you want 16-bit, but I don't recommend that for mixing.)
    • Or alternately, go to File > Export > Export Audio if you want FLAC. (24-bit FLAC is my preferred choice, but I'm weird that way.)
    The difference is not night and day, but it is audible. In fact, it is also visible on a spectrogram. (I checked just to make sure I wasn't imagining things.) Doing the above will reduce the ever-present brittleness as heard in native SynthV 44.1 kHz exports.
    peaches2217
    peaches2217
    Thank you so much!!
    The disappearing/reappearing Synth V Studio transport controls are really getting on my nerves. Mr. Hua has some serious explaining to do.
    Has anyone else noticed 18 kHz glitches in Saki AI? Not that it matters too much, considering that YouTube's compression cuts off audio somewhere around that frequency. But still...
    Welp, here's the opening two minutes of my first Saki AI song:

    This was a lyrics-first song, and りくりくりさん give me more lyrics than I could handle. It's not that there's a lot of lines, but each line is very lengthy and there's no pattern to the syllable count. Oh well, I did what I could to fit the music to his lyrics, and there's even some word painting.

    But the whole thing is a bit like a sumo match—long moments of inactivity followed by a few brief moments of action. The full song is nearly five minutes long, and only the last 20 seconds are interesting. (Not included in the above excerpt.)

    Better luck next time, I suppose...
    Twillby
    Twillby
    I think it's really cool! Though maybe I'm biased because I liked the sumo match comparison :teto_lili:
    MillyAqualine
    MillyAqualine
    OMG So realistic ;A; It fits her vocals, holy, I imagine her being a real human (or even her VA) performing rock on stage (Indochine x Saki AI when ? )
    I
    inactive
    I received a reply from the lyricist, and he also thinks it's pretty cool, mostly because the music (more-or-less) fits the words.

    But he also pointed out that the sections with short, rapid notes remain very clear and understandable. I never really thought about it, but yeah, Saki AI gets through those parts better than many other vocal synth tunings that I've heard. (Of course, this is highly dependent upon who tuned what.)

    Now, do I add the breath effect?
    Does anyone know if it's possible to keep Synth V Studio Pro's transport controls visible at all times? It's really annoying how it keeps disappearing when doing certain functions, despite the available screen space remaining essentially the same.
  • Loading…
  • Loading…
  • Loading…
  • Loading…