KClip 3.5 Testing (versus 3.2.4)
Вставка
- Опубліковано 2 жов 2024
- Some of my viewers said the new version 3.5 of KClip 3 was problematic (on Windows), so I did some testing of 3.2.4 performance vs 3.5 performance. Summary: if you turn off the new visualizer, performance is essentially the same. If you enable the new visualizer, 3.5 uses up 15% more DAW CPU resources than 3.2.4. Also, the new visualizer is worthless. The old visualizer was much more useful for real-world use cases. On the plus side, you can make the KClip window nice and small in v 3.5.
EDIT: Grrrr - OBS flaked on me during the very last part and didn't record the end. With 50 instances and the Visualizer turned OFF, the Load (AVG) (i.e., the "CPU usage") was 0.460. Which was exactly the same as 50 instances of the older 3.2.4 KClip.
CONCLUSION - Turn off the worthless new visualizer and it's the same impact as 3.2.4, and you can at least make the window smaller now. Or just stay on 3.2.4 until Kazrog can improve performance and (hopefully) revert the visualizer back to a more useful "crest factor" view like before.
I agree. I like the old visualizer better.
If the visualizer was as good as psyscope this plugin would be a gem.
#Kazrog Read This.
@@durimanuriplaylist2660 Making a cross platform scope as good as psyscope is a very tall order, guys!
@@KazrogPlugins Cause we knowing you can do it and we got the time to wait! Thank You for the already great Software.
@@KazrogPlugins my money is waiting for you 💸💸💰 💰
Haha thanks for doing this so quickly Baphy! Looks like they're going for a fabfilter look with the new GUI, I like the aesthetics but will probably have that visualizer off because yeah, pretty useless. Going to do similar tests with my setup tomorrow to see what's going wrong in Studio One.
No hard feelings about your opinion, but I’m surprised you don’t find the new visualiser MORE useful. At the top, in red, you see the clipped portion of the waveform, giving you real-time feedback of how you’re affecting transients.
Love your videos, and your methods are very good with one caveat: you should always use oversampling when rendering with clipping. 😎
I wouldn't say it's useless, I think having both to choose from would have been my preference. I did like the previous graph but I do see value in having this one too, but not as a complete replacement of the previous.
maybe we could get the red clipped portion on visualizer that shows the full waveform? I think that would be best!
A main element of his technique and his teaching centers around seeing the audio waveform in an oscilloscope so you can understand how it’s being changed through processing (in this case clipping). The new visualizer doesn’t show enough of the waveform to gain this perspective. The same could probably be said for ProL2 or other visualizers like it. A high resolution oscilloscope would be much more useful for the techniques he’s teaching. Basically the old visualized but with more detail and information. As far as the oversampling, I don’t think he has ever said don’t do that but he does make it clear that it doesn’t always sound “better” or rather give you the desired effect you’re looking for. Not to sound cheesy but I believe the saying “if it sounds good, it is good” would apply nicely here.
@@kagaro99 I find it hard to picture how it would not be better to oversample, most of the time. The problem with clipping is that it will always generate harmonics, higher in level and in how far up they extend the more you clip. Without oversampling, as soon as those harmonics pass the Nyquist frequency, you have aliasing, basically overtones reflected back into the audible range, but are unrelated to the fundamental frequency that generate them. In other words, they don't sound musical, but dissonant and tend to give sounds a harsh and fake sounding character, characteristic of amatourish productions. You can get away with no oversampling on lower frequency sounds and light to moderate clipping, but you always need to judge if aliasing becomes audible. I haven't heard of a good, respectable engineer saying aliasing sounds good or is not an issue.
The waveform was clearer in the previous version due to colour and zoom level. It's useful to have a clipping indicator, but not so much when it makes the waveform harder to see in detail. Adjustable zoom and colours would be fantastic. Excellent plugin in any case!
I just wanted to say thank you so much for the new CTZ series, it's been incredibly helpful. You could and should be charging for this as you've put a huge amount of work in.
For me its the opposite. The new visualiser seems somewhat useful as its more like SIR Standard Clip. The old visualiser was completely useless to me.
Good to see Apple Silicon native support. Thanks Kazrog for making this plug in, even better.
Thanks! By the way, Apple Silicon has been supported for over a year now, but the new update makes things play nicer with macOS Monterey.
@@KazrogPlugins nice one. I also own True Iron and True Dynamics so thanks for silicon native versions of those as well. Also good to finally be able to open the interface without that glitch. That made me stop using Kclip, but now I am going to rinse it again! I also use Standard Clip...with the ceiling on ;)
KClip - Version 3.6.2 - May 1, 2024 is out. would love to hear your opinion which version is actual your favorite.
Congrats on 15k, Baphy!! Great content as always ❤️
It's funny man. I just buy whatever the hell you recommend. Haha. These guys owe you money!!! I bought psyscope pro (my 1st oscilloscope), duck (my 1st LFO tool), and now kclip (upgrade from using gclip lol).
I personally think they did a great job with this plugin. Love it.
And what was the result without visualisation?
Grrrr - OBS flaked on me during the very last part and didn't record the end. With 50 instances and the Visualizer turned OFF, the Load (AVG) (i.e., the "CPU usage") was 0.460. Which was exactly the same as 50 instances of the older 3.2.4 KClip
@@Baphometrix ah OK thanks for the results, was looking for that, Im gonna stick to my version, updating is sometimes worst
@@Baphometrix Have you tried enabling OpenGL in KClip settings?
To me this testing was maybe interesting, but practical expected usage would not use anything close to 50 instances, in my opinion. Maybe some do this, but the resources are minor, so not a concern. On second thought, maybe the light resources was his point. Which he definitely showed. I might be interested in the graph showing clip amount and/or colored intensity graph, but even a numerical display of percentage would be good in my opinion. Just a noticeable reference that can back up our ears, lol.
Video ended early unfortunately
The fifty instances were too much i guess. :D
(just jokin')
they probably would, on my system
Grrrr - OBS flaked on me during the very last part and didn't record the end. With 50 instances and the Visualizer turned OFF, the Load (AVG) (i.e., the "CPU usage") was 0.460. Which was exactly the same as 50 instances of the older 3.2.4 KClip
3.5.1 doesn't show gain reduction anymore? wtf
What happened to the "The problem with oversampled clippers" video? I was watching, went out for a walk with my dog, and when I came back and pressed play it said: This video is private :(
Myself (and others) pointed out that Baphometrix forgot to turn on the ceiling in the competing plugin, StandardClip. The behavior of the ceiling feature in both KClip and StandardClip is the same (KClip was first, though!), and many of my users own both plugins. Christian Knufinke is a great developer, and I think it's important that any competitive comparisons are fair to all products involved.
Lemme tell you what I think has happened. He made a 35 minute video trashing StandartCLIP and praising KClip on the false premise that StandartCLIP allows peaks to exceed 0 when oversampling is enabled, whereas KClip doesn't. The problem was that he didn't notice StandartCLIP also has a "ceiling" function (just like KClip) which needs to be enabled. Some people in the comment section pointed out that he forgot to turn on the ceiling on standartclip but without even checking if they were right, he responded with a simple: "look closer". I guess in the end he noticed his mistake and removed the video.
Be careful criticizing this man or pointing out mistakes in his videos though. He can get very defensive and nasty :). This happened to me couple of times
I misrepresented something about Standard Clip in that version, and UA-cam won't let us upload new edits over old published videos, so I took it down.
@@aysegulozguler7712 You'll notice I'm leaving this comment intact, right? 😉 Yeah, I messed up something in that video. Some folks pointed out the mess-up. I quickly took it down to not propagate bad info. Working on a new version now.
@@Baphometrix Thank you so much for leaving this comment intact. You are a very "merciful" youtuber
I recently purchased kclip and did my own experiment. I don't have a resource monitor like baphy was using, but after setting up my CTZ template and having instances of kclip everywhere, Ableton began crashing on me every time I began working. It took me a while to narrow down that kclip was the actual problem, but I did the same thing baphy did here. With both 3.2.4 and 3.5, multiple instances(like 15 or more, and I run a lot of tracks so that's not that many) of kclip with the visualizer would crash Ableton everytime with or without opengl. With the visualizer turned off I can run dozens of tracks with kclip with literally no discernable effect. The interesting thing is that the CPU meter reflects no difference, but the performance difference is shocking. I don't have the greatest laptop but I have never encountered anything like that. Turn that visualizer off
I have the same issue on Ableton + KClip. All genuinly bought and KClip crashes Ableton, even prevents me from loading the project when audio is active. So if you have the same issue, turning off audio and then deleting all the KClips as the way to go.
Turning on OpenGL fixed the lag I was getting in the older version. This is great, however I agree with you about the visualizer. The old version was way more useful.
can I do that on3.1.0v?
@@mosheshohat3486 I don't know, i'm using version 3.5.0
@@finnhuman1012 I upgraded today so it's OK now I have it
The old visualizer is def better, hope they go back.
I’m surprised anyone found the old visualiser more useful. The new visualiser shows you the clipping amount in real-time.
In any case, more visualiser style options could be added down the road including the old style.
A hi res oscilloscope would be incredible!!! Would only have to use one plugin when clipping!!! Also better metering could be added somewhat like standard clip instead of just a number. The new visualizer doesn’t show numerically how much clipping is actually happening.
@@KazrogPlugins I think people here rely on extra visualiser so they are biased.
But objectively the new visualiser is definitively more useful.
@@KazrogPlugins I mean sure, it shows that clipping is happening, but it doesn't show you how it has affected the shape of the waveform after the fact. The old visualizer showed that clearly, which was very useful. Not tryin to be rude here, because I really appreciate y'all putting out the clipper in the first place, but I am looking forward to the different visualization options down the road
@@KazrogPlugins I like the new one but it should be oscilloscope but stereo I mean, like previously up and down but with LINES showing 0db/clipping, and then spikes. I hope it makes sense
It doesn't matter if kclip is doing any processing or not? Thank you very much for all your work btw - trying to incorporate some of your ideas into ambient music.
Grrrr - OBS flaked on me during the very last part and didn't record the end. With 50 instances and the Visualizer turned OFF, the Load (AVG) (i.e., the "CPU usage") was 0.460. Which was exactly the same as 50 instances of the older 3.2.4 KClip
If you enable OpenGL in the plugin, the visualiser will use the graphics card rather than the CPU to do the visualiser. One caveat: if you're running Windows OS, you will only want to have 1 KClip window open at a time when using OpenGL (for now.)
U got any vids on oversampling clippers?
Lol I know I saw the notification for that video all day and went to watch it and now it's private. Make's me wanna watch it even more.
@@nickthorne5872 I'll shoot another take of that idea tonight, hopefully. I made some mistakes in the first one.
Here for this
@@nickthorne5872 I mean I was just the convo between Kazrog n Baph about it n now I gotta know cuz Kazrog was saying that depending on how u set ur oversampling, the clipped noise will end up in the final render? I thought it was anyways?
@@Baphometrix I’d read the convo below between u n Kazrog about it n now I gotta know cuz Kazrog was saying that depending on how u set ur oversampling, the clipped noise will end up in the final render? I thought it did anyways?
4:18 When she says you messed up her hair
Reminds me of Flatline.
Love all your videos and the way you explain everything, i love the oscilloscope you use, i normally use Oszillos Mega Scope by schulz audio im mbeat mode for better detail, would be amazing to see it in your tutorials
Can confirm - just bought it because I'm implementing your strategy and the UI on the new version still broken. Rolling back to 3.2.4. Thank you.
2 questions :
1) if i were to purchase k clip , do i have the option for the older version ?
2) Im currently on episode 8 and my question has to do with your starting point drum template foundation and how you get you desired loudness from that starting point . you went over the drum buss processing but i don't think you covered the individual sound processing for the drums. is it the same approach you demonstrated with the rolling bass synth and other examples using the kclip and dp meter pair method ? is there an episode where you go over that ?
There is a LOT to sound design for drums. Especially loud drums. Especially loud drums that won't sound like a rubbery cheese-grater crunchy in your typical club system with really bad acoustics. So... maybe someday? But it's a deep rabbithole to dive down.
As for older KClip vs the new 3.5 KClip, IMO just get the new one. It's perfectly fine. It's what I use now.
I was just gonna ask you about the new Kclip update and boom you already have a video, you are the man !!!
Personally I have very bad experiences with updates messing up things so I always wait to see peoples reactions or patches from the companies, the new look is horrible , at the last run the video gets cut so it doesnt show if turning off graphics helps
Grrrr - OBS flaked on me during the very last part and didn't record the end. With 50 instances and the Visualizer turned OFF, the Load (AVG) (i.e., the "CPU usage") was 0.460. Which was exactly the same as 50 instances of the older 3.2.4 KClip
Sorry to hear you don't like the new look! I'd like to add some selection of skins/themes in the future, as well as more visualiser options, but it was time for a new look and a better workflow, and I'm very happy with how KClip 3.5 came out, as are most users.
@@KazrogPlugins thanks for replying, its not easy to please everybody 😉😂 You have the best Clipper of the market, its a good option to have different skins
@@KazrogPlugins An amazing clipper. I just bought the last version. I wish it will get better graphics as well and Zero latency also will be amazing! thanks, amazing job so far :)
I was really excited to see the new update with a big jump
but I was a little disappointed with the visual.. It's as if they have gone down the Flatline or fabfilter look. But they had their own look originally and now made them self look like a copycat.
Thanks for your feedback! There's been no effort here to copy anyone, and KClip was the first cross platform mastering clipper plugin on the market.
The color palette in KClip 3.5 is different than either plugin you mentioned, although I will say that the workflow itself has been inspired a bit by FabFilter. Some more options as far as theme and visualiser style are coming, though!
@@KazrogPlugins I know you guys are not copycats. I said you have made yourself look like copycats as the GUI now looks like Flatline
I've emailed you with greater detail to my thoughts
Yeah the new visualizer sucks imo..I already emailed them lol. ask if they add the old one back in and just be able to switch between the 2
Sorry to hear you feel that way! The majority of our users were asking for exactly the type of visualiser introduced in the new version 3.5. However, I can understand that it's impossible to please everyone with only one setting, so the old mode will be coming back as an option. Cheers!
@@KazrogPlugins Thanks for the quick response, KClip is one of my favorite plugins. I appreciate the Kazrog's concern and kindness for its users.Thanks Again
I agree. I had the same thought. Having both would be perfect. And I would USE them both in different ways, unlike the old spectogram GUI which I found no use for. The original waveform GUI is so useful to gauge the overall height/crest factor of any given waveform being clipped. It's just useful information to have. It's a shame to see that feature go.
@@nickthorne5872 They said in the next update i would be added back in ..So now we wait lol
@@theweaz9975 Heeeck YESS!! The perfect clipper plugin awaits!... Cool :)
Soon as I got the email, I downloaded and tested it.
I actually find the visualiser useful. I can use both my eyes and ears while on one window.
Thanks!
Oh and the new Kazrog GUI is crap. The old was way nicer
Is kclip better than NF Saturate?
Saturate is a bit different from KClip.
Yes. And No. The answer is complicated. KClip is absolutely better to use at the start of every project, and all during your early composition/arrangement and early mixdown phases. Only at final mixdown might it be advantageous to replace **some** instances of KClip with either NF Saturate OR with TrackLimit (from DMG Audio). I'll be doing an upcoming "Mixing Loud with CTZ" episode talking about when/why to **maybe** replace KClip with these other two.
@@Baphometrixmore transparent to me.Plus cheaper with edu discount or bf deal
How to handle Loops in CTZ Strategy ?
I don't understand the question? Can you elaborate?
@@Baphometrix loops Mostly Not following CTZ Strategy,What we Should Do when Facing This kind of Sounds in our mix ?
@@Baphometrix also On Live instruments recorded in Studio
@@sahandmz4212 You neglected to elaborate on your first question in your second comment. The second comment was even more inconclusive.
@@sahandmz4212 Okay, I think I understand. You mean when you pull a "loop" from your sample library. Something that has multiple different instruments/sounds baked into the same single sample. Well, those are ALWAYS problematic in that they are pre-mixed chunks of content. And therefore you have no control over re-mixing them to fit the context of your overall project better. When you use a loop, you simply have to make it "work" in the context of your project, or reject it because you cannot make it work. So it's the same with any type of dynamic range control problem you might have with the loop. If you can saturate it and/or clip it to get it loud enough to fit into your mix, and it still sounds good after that treatment, then great. If you can't get it loud enough, or it sounds bad after you do what's necessary to make it loud enough, then you need to throw it away and find a different loop instead.