RexRed

On 7/31/2023 at 11:37 AM, JohnnyV said:

Interesting the Zoom uses 32!  I think that is what Ulf was getting at. 48/24 is about as good as it needs to be considering what music ends up being played back on these days it is really overkill. 

I have noticed a huge jump in sound fidelity and recording quality when switching to 32 bit. Quality which incidentally gets obliterated with the application of Melodyne to my clips.

Regardless of what bitrates people are listening to music with I would still prefer my own masters to be the best.

And with the utmost respect, I disagree with Ulf and Melodyne sounds crappy how it handles many sibilants. Someone has got to say it.

Is this Melodyne or the ARA specification? I think the latter, but both are inexplicably linked.

I will be recording all of my future songs in 32bit.

It is hard to believe that a $200 Zoom box can beat a $1,300 RME Fireface UCX 2 with recording but that is indeed the case.

I still run my Zoom into my UCX2 and use my UCX2 but for recording it is Zoom all the way.

It is weird but I think Cakewalk runs more stable in 32 bit mode as well.

On 7/31/2023 at 11:06 AM, JohnnyV said:

Melodyne uses a totally different rendering system and it will be 24 bit. 

How do you know Melodyne (ARA2) is 24 bit? Please provide documentation.

My offer still stands 200 bucks plus shipping for 2x practically unused  Behringer U-Phoria UMC1820 USB Audio Interface. ( I paid 600 for the pair)

Perhaps the worst purchase I have ever made in my life...

Hey awesome Cakewalk peeps!

Here is my latest song I made about two cowboys who fall in love.

Love is love peeps!

😊❤️👍

1 minute ago, Glenn Stanton said:

nope, it's definitely not ASIO4ALL. 🙂 its a real ASIO driver which works most excellently.

I beg to differ. It says ASIO but it is not.

3 minutes ago, Glenn Stanton said:

@RexRed @Starship Krupa the lower end units use the ASIO4ALL, the HD and larger units have real ASIO drivers - one of the reasons i picked the UMC202HD - it uses the same ASIO driver as my larger UMC1820 unit.

as a note: i have ASIO4ALL installed but don't use it EXCEPT for a couple of older VST synths via Tone2 where is seems to work the best of all drivers. no idea why, but it just does. so it's hanging around solely for that. otherwise everything else is either the default system drivers or disabled 🙂 

The 1820 is not a true ASIO driver (so I have been told) it is a rebranded ASIO4All driver dressed up to look like an ASIO driver.

33 minutes ago, Starship Krupa said:

That is only true for the very bottom of the line stuff. Once you take a step up from the bottom, they do have "real" Asio drivers.

Another case of Behringer shooting themselves in the foot by further messing up their reputation.

Yea, I bought two of them, only to find out they are stereo output only and not true ASIO.

https://www.sweetwater.com/store/detail/UMC1820--behringer-u-phoria-umc1820-usb-audio-interface

They both sit in a junk drawer.

They are brand new I would sell the pair for $200 plus shipping any takers?

23 hours ago, Glenn Stanton said:

so your saying that the disabled driver got re-enabled?  i just leave my devices alone except for the realtek.

and just use my Behringer ASIO drivers.

Someone once told me that Behringer's ASIO drivers are not actually ASIO but a rebranded form of ASIO for All.

 

On 7/28/2023 at 5:48 PM, Glenn Stanton said:

you need to disable the driver in device manager, not uninstall as windows will see the hardware and load the last known driver. if you disable, it will remain disabled (unless some update reactivate it) and it won't keep reappearing. same for the realtek etc any drivers except ASIO (which can be uninstalled usually, if not, disable it).

Disabling Nvidia drivers in device manager is the way to simplify your sound options although, whenever you install a new Nvidia driver update from GeForce Experience it will reenable them so you have to go back in to device manager and disable them again.

One would think that the new Nvidia driver would recall the previous state that they were disabled but that is currently not the case.

Realtek integrated sound can be disabled in the bios in most cases.

4 hours ago, David Pollock said:

My very basic Nektar Impact GX61 keyboard is causing softsynths to play back a half step higher at random times in multiple projects. Disconnecting and restarting a project makes that issue disappear, and plugging the keyboard back in will eventually result in the problem again. It obviously makes it nearly impossible to work on anything.

I know it's a long shot, but I am wondering if there is anything at all I might can do to fix the issue without replacing the keyboard. The back story is that I have a very nice keyboard in storage that I cannot get to right now. So buying another one is something I don't want to do if I can help it. I did a factory reset on the keyboard. Beyond that, I don't know of anything else to try. Any suggestions?

I don't know if my issue is the same but I have an M-Audio Oxygen 49 keyboard that I love.

My pitch bend went erratic and floating above and below zero making it impossible to use.

I tried to repair the synth and nothing seemed possible other than buying a new one.

Instead of buying a new one I put it in storage and bought a cheap controller instead.

One day I pulled the M-Audio out and it worked perfectly fine.

Either something got into the keyboards firmware and got reset with time off, or a Windows or Cakewalk update fixed it.

I got a Nektar SE 49 keyboard that I don't like the action but it works okay.

Note: You have to make sure that you set the channel that you are active with midi on and set the device as Nektar "Omni" and then change the midi  channel to "channel 1".

Otherwise you can get errant midi from other tracks that bleed in. 

Months went by and a Windows update happened and suddenly my M-Audio keyboard is back to freaking out with pitch bend again but a slightly different kind if error but still there regardless. I am back to using the Nektar. I suspect it will work again at some point.

It is nice to have other channels communicating midi because Cakewalk is able to record from complex VST instruments on multiple channels, banks, patches etc... 

But having these lines of communication always open from other tracks can also bring unwanted data in.  

1 hour ago, DaveT2 said:

Very nice tune.

Thanks a lot Dave for the nice comment! 👍😊📻🎼🎙️

I would like to see ARA 3 with 32-bit support and colored blobs with track color information passed on to Melodyne per track.

Orange blobs all stacked on top of each other is messy as heck.

Many have requested colored blobs per track to Celemony support and they have so far opted not to acquiesce. 

I requested colored blobs/track 4 or 5 years ago.

The grid lines are also hard to see in Melodyne sometimes even with high contrast set all the way up.

Another feature request that has failed to materialize.

1 hour ago, jack c. said:

what is the vst plugin guitar???????????jack c

The guitar is a couple of samples joined together from splice.com

Then I used Scaler 2 to figure out the chords for the pads I added.

The chords were so complex that I did not think I could try and play them.

Scaler 2 did a fantastic job making midi chords of the wave samples.  Thanks for the compliment on my song Jack!

🎵🎹🎺🎧

My new song, vocal from mostly 1 take, no Melodyne or tuning applied, new 32 bit audio interface.

Yup, 32 bit vocal recording in Cakewalk and rendered down to mp3 16-bit, 44khz.

Let me know your thoughts awesome Cakewalk peeps!

😊

On 7/29/2023 at 8:14 AM, scook said:

ProChannel is based on the VST2

Perhaps, the biggest improvement of the VST3 plug-in is that it doesn't waste CPU resources and only works when it detects the presence of an audio signal, unlike VST2, which remains active at all times. For users, this means an opportunity to use a bigger number of plug-ins without overloading the system.

Google

https://blacktidemusic.com/software-and-plugins/vst2-vs-vst3/

Another article that does not mention bitrates or conversions of these interfaces.

Melodyne definitely resamples the wave, this is the reason why it works on a 32 bit file in my project.

What bit it resamples the wave to is still unknown. 

Considering ProChannel is on all the time this could account for CPU usage and load balancing may address ProChannel's use on only one core.

And just because a plugin is VST3, this does not necessarily mean that infrequent use is implemented in that particular plugin.

Then we have multi function plugins like Izotope Ozone 10 that seem to rely on many inputs.

It would be logical to assume that no conversion is happening when a VST plugin is used.

But ARA definitely converts/resamples the wave file.

I recall the old ARA that left a ghost file of the original wave that was muted.

And Melodyne must be resampling the wave to a lower bit depth if it cannot work on 32bit waves yet it does seem to work on them.

Does it resample the wave to as low as 16 bit? Still unknown. 16 bit is not out of the realm of possibility.

https://en.wikipedia.org/wiki/Audio_Random_Access

In this whole Wikipedia article and all the links at the bottom, there is no mention, not even a whisper, as to what wave conversion process is going on under the hood with ARA2.

The DAWS are evidently passing the wave files off to Melodyne and Melodyne is resampling them.

Am I sending off to the music stores 24-bit 96khz song files with 16-bit resampled vocals?

And what about effects? Are some VST effects forcing my wave files down to lower bitrates?

VST2 vs VST3?

VST2 VST3 and ARA ARA2 are interfaces.

The question becomes the quality of these interfaces.

Another question, does Pro Channel also rely on VST?

3 hours ago, scook said:

I do not know if the ARA path is 32 or 64bit. That would be a question for the devs.

I posted the Reference Guide pdf reference in case the web-based help failed. The pdf is a snapshot of the web-based help..

There is a link to the pdf at the top of this forum

 

I am guessing the ARA interface is 16 bit. This would explain a lot.

I found the PDF file as well, thanks for the heads up on that.

I asked if variable bitrates would not help with the bends of notes/blobs.

My email from Celemony

************************

Hello Rex,

thanks for the mail. Melodyne does not support 32 bit recording and variable bitrates. I doubt it is in any case halted to transients missing.

The bitrate sets the resolution of the amplitude at a given time. I do not know of any recording software that would assign bit in that way.

 

With best wishes,

Ulf Kaiser

*************************

Ulf has been a great help every time I have had to reach out to him.

In the next email he sent me links to some tutorials for Melodyne.

They were very helpful but it became clear that Melodyne is really only supposed to be applied to very specific parts of notes. Namely, the part where the main tune is, omitting the drift up and from the note/blob. 

Therein lies the rub, when you apply Melodyne to a section of the blob where the pitch "bends" this is where you get the wonky sound.

It distorts the tone in a boxy and unflattering way.

So it is either, get your bends right or get your bends right.

Melodyne won't help much on those areas.

Then I got this email message.

**************************

Hello Rex,

there's no need for 32 Bit recording or 96 kHz. A proper voice recording will do fine. 

The rest needs to be done using the tools in Melodyne: Separation, Note Assignment and Sibilance etc.

Make sure to watch a few videos here, it might help: 

See: https://helpcenter.celemony.com/M5/doc/melodyneStudio5Training/en/M5tour_Videoseite?env=proTools

Producing skills: Lead vocals

https://helpcenter.celemony.com/M5/doc/melodyneStudio5Training/en/M5tour_Videoseite?env=proTools

 

With best wishes,

Ulf Kaiser

************************

While these video helped a lot it also highlighted the fact that many parts of your recording cannot be altered with Melodyne. 

This brings me back to the question. If ARA 2 is only 16 bit conversion this might explain this limitation.

Ulf says Melodyne does not support 32 bit recording. 

What he does not say is whether if ARA 2 supports even 24 bit recording.

As for needing 32 bit Support in Melodyne I can see how the tools in Melodyne can be utilized to a great extent to mitigate the tone corruption of blobs. But I wonder if having Melodyne applying itself to higher quality wave files would not also be beneficial.

I am wondering if signing my lines in key is not simply the best option.

 

38 minutes ago, scook said:

half way down this page https://legacy.cakewalk.com/Documentation?product=Cakewalk&language=3&help=Mixing.27.html

pdf p1013

I am not sure this worked right.

It took me to a help file about, "How to use real-time effects".

I can see how Melodyne is sort of a real-time effect.

I think this is using effects on the clips bin.

How do I find pdf p1013? 

18 minutes ago, scook said:

I was adding that to my reply

 

Internally Cakewalk is running 64bit because you have the 64bit DPE enabled.

Plug-in may processing at one bit depth but communicate at another.

The DAW tells you the bit depth of the connection to the plug-ins. This and the 64bit DPE determine the bit depth between the DAW and plug-ins like Melodyne.

I am looking for the link just a minute

Thanks for the info SCook.

A link would be awesome. 👍

When I right click on the audio clip and select associated audio files it says, bit depth 32.

26 minutes ago, scook said:

The value shown in the Transport module is NOT the Audio Driver Bit Depth.

The value reflects the Record Bit Depth found in Preference > File > Audio Data.

 

 

Okay, I found that, I know I have been told that before but I forgot it.

Why is Melodyne working on this 32bit wave file?

Is it converting it to 24 bit?

I would think Melodyne would give me an error and say, "This is a 32bit file".  

Is Cakewalk playing both 24 and 32 bit files in the project?

I start a new project and record a segment and the project says 32 bit, then I close and reopen the project and

up in the play transport under the now time it has changed to 24 bit.

But my interface says it is running at 32-bit float

The preferences say 32 bit.

Then when I go to put Melodyne on the recorded 32-bit file it works.

Melodyne is not supposed to work with 32bit files.

What is happening here?

Is Melodyne converting the file to 24bit?

Are 24 and 32bit files both working in the same project?

 

 

 

7 minutes ago, Byron Dickens said:

Irrelevant.

 

That's certainly enough time to get practiced up....

 

I'll put money on there being nothing at all wrong with your singing voice and you're just overly critical and have unrealistic expectations of an inhuman level of perfection.

Experience is certainly relevant.

I just like the idea that if I need pitch correction on a certain syllable that it actually works and does not produce a wonky tone and not work.

Having tools that work only part of the time is not the best way to accomplish things.

When those tools actually degrade your work then it is worth trying to improve the quality of those tools. 

20 minutes ago, Glenn Stanton said:

thing is - strawberry fields wasn't about pitch correction on a single part, but rather some clever matching to get the two takes - done in different keys and tempos - close enough to become one. but for as long as tape has been around, people have been using variable speed (when available) to adjust performances - like a singer needing a slightly lower key to sing to, and then sped back up etc. it was actually fairly common in pop to speed up material after it was mixed to increase the tempo and key - more excitement and of course fitting into the 2:30 time slot for songs in order to get air play...

I think they called that "more excitement" back in the day, "hype". 

22 minutes ago, Byron Dickens said:

 

Reductio ad Absurdum

 

Totally bogus examples because altering pitch and tempo through varying tape speed doesn't concern fixing substandard performances.

 

Critical thinking is a valuable skill.

If that was the case they would have used only one take...

Yes, critical thinking and rudimentary deduction is required. 

Do you even have any experience recording with a reel-to-reel?

I recorded about 200 songs on 8 track reel-to-reel in my younger days. 

"Reductio ad Absurdum" aptly describes your criticism with pitch correction. 

 

6 minutes ago, Byron Dickens said:

Bullshit.

That response perfectly matches your knowledge of this subject.

Excerpt

Variable tape speed

By the 1940s, studio engineers could produce primitive pitch correction by tweaking a reel-to-reel magnetic tape recorders varispeed. This process became more popular in recording studios during the 1950s and 1960s. By slowing down or speeding up a part of a recording and splicing with the tape containing the majority of the song, engineers could alter pitch. Another method of varispeed pitch correction was to slow a tape machine down, re-record a new part at a lower pitch, and then bring the recording back up to its original speed.

One of the most famous examples of varispeed pitch correction is the recording of The Beatles 1967 single “Strawberry Fields Forever.” The song as we know it is not one single take, but two takes (7 and 26) spliced together by producer George Martin and his innovative engineer Geoff Emerick. The Beatles had recorded these two takes at slightly different tempos and pitches. By speeding up the slower of the two tracks, Emerick was able to match the tempo and pitch.

https://www.izotope.com/en/learn/a-brief-history-of-pitch-correction-in-music.html

It would be nice to have a producer or two and and few "innovative engineers" helping me make my music.