I'm just going to be editing using card readers...where did I put the boxes of cards now??

http://www.reddit.com/r/AskReddit/comments/rmb3r/how_big_wou ld_an_i5_processor_be_if_instead_of/
Bumblebee ASIO?

I sort of thought this thread might be about how efficient the hosts were just within themselves. Which host hosts better is obviously up for debate, and apparently difficult to even measure, but how they work internally with just their own code is something that would be interesting to test. Like, tiny tiny differences in handling simple midi, simple routing, simple audio playback.

It is only logical to me that even basic things would be built differently because different hosts were built in different ways with various focuses.
e.g. One team made audio a priority, one team midi, one team plugins, etc..
Of course they are all 'full featured' hosts but they probably had differences in their underlying ideas. I guess this wouldnt be a fair comparison because not all hosts claim to be the end all in all areas, but it would be interesting.

I don't know, I've seen some pretty hefty commercial beasts that are optimized pretty darn well. I think you'll find that many times a commercial team will have more time and resources to put into optimization, including developing their own libraries, etc.

And of course, there are plenty of freeware Synthedit creations that are a huge beast because they are built out of stock modules and the creator doesn't even have the knowledge to optimize it.

Not ALL the time, of course. Just pointing out how it can go both ways. I think there is zero link between commercial and CPU usage.

Brent

Once again: i wasnt saying that commercial = unoptimized.

Only: commercial = more features = more CPU cycles consumed.

not true either

More features = more complex GUI = more choice for the user = less choice made by the programmer = nothing about the CPU usage.
Of course a feature can be a hidden one too, but to an extent, speed optimization is a feature as well.

I mean I shouldn't care, I do commercial plugins so it's nice to see that I'm allowed to eat more CPU than freeware.. that's too bad for those who make freeware I guess
What's also sad is that this kind of thought will only motivate plugin makers not to optimize their code, for the reason that "if it eats a lot of CPU, the user will think it sounds better". & this is how you end up with overpriced plugins with oversized GUIs that are also CPU hogs.


(I will also add that the size of the GUI has nothing to do with the CPU usage of the DSP either.. because some people expect "a big commercial plugin" to be allowed to eat more CPU because it has an oversized GUI, thus to be as heavy on the CPU as it looks)

IDK how can you misunderstand everything i write here.

Looks like you actually want to.


Can you tell me what will take more CPU cycles ?

A. An optimized simple oscillator that doesn't emulate all the details of the real thing accurately (found on most freeware).
B. An optimized accurate emulation of an analog oscillator (found on few best payware).

or a fish?

Obviously i meant features used constantly and all the time - features that you can't turn off.

Like for example free running oscillators with analog drift and other inaccuracies modelled (every cycle unique).

that sort of thing actually isn't a huge expense, although you're right that it can be.

where you're really wrong though is the idea that only "commercial" products would have such features. not true at all. in fact free products are far more likely to be impractical cpu-hogs because they don't need to sell.

i know we see an awful lot of "beta payware" but honestly, it should be easier for freeware to conduct experimentation on the public.

see some of the stuff xoxos writes. a lot of it there is no way anybody would ever use that in a hit club track and pay 399 euros for it. it's definitely unique and experimental though.

you also have to keep in mind who's seen source code and who hasn't

I'm going to write a sequencer in PDP-8 FORTRAN and record/play to/from DECtape!!!



I'm still finding I can get 2 more instances of Alchemy in Reaper than Orion 7.6, but CPU isn't my issue so much at that point as RAM.

I need one done in COBOL...think I'll dig out me one book on it...

As for the RAM issue...I now have 16Gb of it..going to 32 next...

I didnt say anything like that
Care to quote me saying what you could interpret as me thinking that "only" commercial synths have such features ? It would be nice to improve my english language conversation skills
Most probably you were under an impression of what Tony thought i said.

---

My last try to make you all understand me correctly:

Take 10 most popular commercial synths made in 2011 and 2012.
Take 10 most popular freeware synths made in 2011 and 2012.
Put your head on a stake and make a bet which group of synths uses more sophisticated algorithms and/or even just more old standard and well known to all tricks to get closer to that pure hardware analog sound.

Great!! Now we are going to start yapping about which VST is going to be the uber analogest synth

I thought this thread would be gone by now, Reaper is known as the best host even prior to this poll (been told so). I guess the trumpet blowing will continue as the flag is raised in victory for all to see.

When will all who doubt this stunning defeat just realize their DAW suck's and they are idiot's for not choosing Reaper in the first place.

If that were the case it would be awful quiet around here.

It's awesome!



It's an absolute lesson in losing the plot completely