display buffer based paint routines are super laggy suddenly with latest commit here as well (also in exported plugin) - had to roll back to a january commit - I also built very latest commit without any of my modifications from the source and same problem. If anybody finds the last working commit so we can find out what's going on that would be awesome :)

@d-healey partially agree on the freebie thing - as the tracking became worse and worse over time for meta internally the platform doesn't know a lot about the purchase behavior of a lot of users anymore - especially iPhone users who often statistically have a higher purchase value. A really good and valuable free product will lead to very low CPMs while still attributing for conversion - maybe 10 times lower than a regular campaign. Means you reach 10 times more people that are tagged as converterters for the same money. You get maybe a cost per result of 20-30cents.
Paid ads are auction based so it's not too much only about - is this person likely to spend money at all - it's more like - is this person likely to spend money RIGHT NOW. So you are still catching often the right people at the "wrong" time. If you have upsells in place a certain percentage will opt in for it - if you are doing it well it will finance the campaign directly. Yes a lot of them will never spend money BUT now that you have their email you can exclude them from the campaign and prevent Facebook showing them ads over and over again wasting your budget. Also somebody that is not ready to buy now (maybe young - no funds) might become valuable later as he has a better job whatever and you would have to pay quite some money to acquire him now via paid ads.

@bendurso I would not think too much about the Facebook roas and look much more at the overall roas of your business. Treat META as a new customer acquisition tool and look at the customer life time value. Example: you have a low cost product you advertise on META for 15bucks - you make no direct profit on it and have a ROAS of 1 on it in a 7day window. every 5th of the low cost purchases buys a 100 dollar product later through email marketing and eventually buys another 100 dollar product a few months later you made an average of $55 per customer. That effectively means a ROAS of around 3,66 for that campaign. Always exclude your customers from targeting btw. You don't wanna spend money on showing your ads to people all the time that you already have the email.

@Lindon demystified version: if you have a good product and good ads (it probably needs some testing and investment to figure out what works) and a good website the algorithm will find people who buy. There are no special hacks or tricks really - most common point of failure (if product and ads are good) is overcomplicating the setup and being too impatient (doing too much changes).

For everything under 1000 ad spend per day on meta for plugins:

Focus on ONE plugin only - your best and most popular one.

One campaign - Manual Configuration - Budget on CAMPAIGN Level - One AdSet per angle/hook - 3-5 variations as ads.

Try starting with 3-4 different angles(ad sets) depending on overall budget (1.5 average order value per day per ad set minimum)

do not touch for one week. Launch new ad sets each week.
The algorithm will def figure out where to spend the money.
Each week turn off ad sets / ads that are underperforming.

If it's going great you can simply raise the overall budget.

Meta will always overreport conversion. Do a custom landing page with a unique url that is not even indexed on google that you only use for your meta campaign.

Each conversion that came from this page or was a direct result from the ads.

If you have a rather fresh pixel (not a lot money spent on) but have an email list, upload it to meta and create a 1% lookalike audience.

Advantage+ is good for scaling.

You have to invest until you have a winning creative. You will find it by testing, analyzing and unfortunately burning some money.

Alway have an eye on soft metrics like CTR and CPM. If CPM is low - people like the creative - if CTR is high - people are interested. I CTR and CPM are good but sales aren't figure out where you are losing them.

@Christoph-Hart ok i managed to compile it - and if this is what you have heard I understand now why you are not impressed. I downloaded the vst and tested it and it sounds pretty different and has also a completely different performance. I'll do some digging.

@Christoph-Hart one last thing - what is the magic trick to make it find the files referencing each other?

'../stftPitchShift-main/cpp\StftPitchShift/RFFT.h' file not found

#include "../stftPitchShift-main/cpp\StftPitchShift/RFFT.h"

@Christoph-Hart haha thank you - I will have a completely non-judgmental look :)

@Christoph-Hart can you share the node template with me? I'm doing something wrong - I get 2000% cpu even though I'm pretty confident I do the processing in buffers haha

btw in general It would be interesting to see how you write a wrapper around a juce plugin code and integrate it as node :)

@Christoph-Hart can you point me in the direction how you included the library? i fail drastically somehow.

@Christoph-Hart did you align the pitch shift with the timbre shift and set a timbre frequency to around 2.5 - 3ms?

if the formants are not shifted together with the signal it will sound strange :)

@Lindon @Christoph-Hart
stftPitchShift is def the best out there I could find - it can do polyphonic and also formant shift.
2048samples latency is normal for pitch shifting audio (it's also normal that low signals sound bad when pitch shifted in those kind of plugins)
I'd love to see that implemented as well :)

@Lindon @Christoph-Hart after posting this I started to write my own c++ code inspired by all the different algorithms available and can say that for "realtime" pitch shifting there is not really a one fits all solution but it's very doable and I got it up and running (but still optimising) . my use case right now is vocal manipulation and besides the fact that you have to really play around with window sizes, overlapping and so on - there is some other stuff that is needed for a industry standard result. I saw that autotune or little alterboy are reporting very odd latency to the host. so I figured there has to be additional sfft processing happening and it is resonance suppression. so before doing the actual shift I lower the magnitude of bins (above a certain frequency) to avoid resonances hitting the shifting algorithm. formant shifting sounds shit yet but I'm working on this right now.
Summary is that I don't see myself coming up with a node that I just use in different projects.
and if you go down the road, don't accidentally process per Frame or your computer will explode ;)

@Christoph-Hart 1024 samples (50ms) is what all the "realtime" pitch shifter plugins introduce.

the signal smith demo https://signalsmith-audio.co.uk/code/stretch/demo/ set to 50ms block size sounds on monophonic signals pretty much how you would expect "industry standard" rt pitch shifter to sound like .
would it be a crazy amount of work to basically replicate the web demo but with an audio stream?

the following looks even better for "realtime" processing and has a MIT license:

used in here:
https://github.com/jurihock/stftPitchShiftPlugin

I tried to implement it already but failed (I think because it's not c++11)

do you want to take a look? I think it opens up really cool currently not possible opportunities straight out of the box.

@Christoph-Hart Oh boy. this was probably the biggest HISE mystery for me in the past year.
Is there any way to work around this so we can have midi out working with all buffer sizes without affecting the performance of the plugin?

@Christoph-Hart ok it obviously worked perfectly fine in a clean new small project. Still it's pretty strange.. I think I deleted a xml network I initially compiled a Faust node in. I still could export the c++ nodes if there was no rnbo node present in the src. I couldn't find a trace anywhere of the deleted xml network but still somehow it's messed up and won't compile even if I delete the whole network binaries folder. is there any way to compile faust nodes independent of script networks like with rnbo nodes. it's super handy especially if you have nodes you often use in different projects.

@ForeverLive Hey there. Thanks for the answer. but I don't fully understand. How do you compile networks separately? Isn't there always one library for all compiled networks per project?

Hey there,
I have some RNBO and Faust stuff I want to use in one project. I can compile my dll networks fine if I either use Faust OR RNBO nodes. I compiled the Faust nodes without problem then created the RNBO patch then tried to compile the whole network again and suddenly the Faust nodes are throwing the following error and the build fails. If I remove all the RNBO stuff out of the src it works again but there is no way it compiles the Faust nodes once the RNBO stuff is in the src folder.

The errors look like that (with different numbers)

expected unqualified-id
float fTemp3 = std::floor(fTemp2);

Any ideas?

EDIT: if the generated .h RNBO Template file is in there makes the difference the rest doesn't matter.

@Christoph-Hart it worked! can't believe it was that simple :)
Probably some valuable info for anybody using the table in scriptnode for more complex timings. thank you

@Christoph-Hart perfect. makes sense!