(positive thread) The RS team deserves praise for the 10.3 performance update.

[chaosroyale]

I know it's a bit late to say so now, but basically, they did a really good job back in 10.3 with the variable batch rendering, and the general overhaul of the under-the-hood performance. Not only for VSTs, it makes a huge difference to Reasons performance generally.

I know I am one of the harshest critics of RS these days, but I with all my criticisms recently, I felt like I was not giving credit where it is due. Their team did a great job with the rendering re-write, and if that had been the "lead" feature for R10, along with Europa and Grain, I feel like maybe they would have gotten a lot more positive response for it. On the other hand, maybe most people don't care about things they cant see?

For anyone struggling with heavy projects on earlier versions, even if you don't use VSTs I think you will see a significant performance upgrade going to R11. The curved automation would be a nice bonus for you too.

See, I can say nice things about Reason.

[bxbrkrz]

[AuroraNovalis]

Is that the performance update enabled and disabled by the "Render audio using audio card buffer size" option? I disabled it after I read it has a detrimental effect on CV wiring.

[Jackjackdaw]

I get excellent performance and stability from Reason. It is the most important thing they get right.

I think v11 is a solid update, I use pretty much everything they added. I would like a juicy 11.5 update before we get taxed again though!

[guitfnky]

yes, that was an awesome update. it's silly how well-optimized Reason is. I regularly have many, many instruments and many plugins on each instrument in Reason, but if I try to do the same thing in, say Digital Performer, things start to get a bit clicky after only a handful of tracks, using the same plugins (and RRP). Reason isn't without its faults, but stability and (now) optimization aren't really among them.

[chaosroyale]

ymmv, but here is my 2 cents:

1: I find the performance in "normal" mode is still slightly better than R9.5 Maybe its just my feeling? OR maybe they streamlined some of the code outside of the buffer size issue too.

2: "detrimental effect". Depends very much on exactly what you are doing. I can tell you about my case. When making large projects I use the larger buffer for the lions share of the composition because it allows me to work quickly and without interruptions. Then when I want to focus on fine-tuning the CV or feedback loops (if you do not use them, this is a moot point) I will switch to the smallest buffer size I can that sounds good and gives the effect I want. I do this nearer the end of the project. If the project starts to stutter I will solo only the stuff I really need and fine-tune the CV, then switch back to the larger buffer to check the whole thing, then render finally at the buffer size that sounded best. One up side I found is that the performance upgrade was so big, I actually need to use less CV to "workaround" heavy VST projects, and can do the job with more CPU-hungry VSTs. For a very CV-dependent track, this would not be a great solution, but these days I am using less CV and more modulation in-instrument, so it has been great for me.

Is that the performance update enabled and disabled by the "Render audio using audio card buffer size" option? I disabled it after I read it has a detrimental effect on CV wiring.

[Boombastix]

They left too many bugs in 10.4 when they moved on to 11, not cool...

[MrFigg]

They left too many bugs in 10.4 when they moved on to 11, not cool...

Agree

[Oquasec]

Yeah I wouldn't mind an 11.5 before the sendoff tbh.
I think we might be near that point

[jam-s]

11.5 would be the right version number for a HiRes GUI rewrite.

[dioxide]

I guess it helps the platform a lot as there are a lot of people who want to use VSTs. It kind of screws up the whole CV thing though.

[EdGrip]

I was trying to use some smooth (but steep) CV for a particular task the other day, and found that I was thwarted by the combinator. For what I was doing, I had to use CV indirectly: CV -> combi programmer CV in, and then direct that programmer slot at the device parameter I wanted to affect. I found this was veeeeeery steppy (only 100 values) and I confirmed it by comparing the direct CV signal Vs the combi-processed CV signal on Skope.

My point is this is the only time I've ever been able to make CV audibly steppy.

[Le Boeuf]

Are you also saying that there is a Big performance upgrade from 10.3 to 11 or am I just wishing?

[Jackjackdaw]

I guess it helps the platform a lot as there are a lot of people who want to use VSTs. It kind of screws up the whole CV thing though.

I only started using VST after the performance update so I dont know what it was like using VSTs before. All the Arturia stuff has loads of modulation built in so I don't need to use CV when I'm using VST and haven't had any issues.

[mcatalao]

ymmv, but here is my 2 cents:

1: I find the performance in "normal" mode is still slightly better than R9.5 Maybe its just my feeling? OR maybe they streamlined some of the code outside of the buffer size issue too.

I remember that was talked on blogs and forums, there was a slight improvement on core and Re devices in 10.3/4. It is not as big but it's a big improvement and it will be noticeable at any setting because the buffer setting does not affect Re performance that always works with a 64 samples chunk. It also depends on Re. I remember testing some effects re's who had a big improvement, and other wich had almost none. But I remember softube's dynamite and antidote had very big improvements ove 10.2.

Sadly 10.3 introduced an error with my controllers so I couldn't use 10.4 for mixing so I stick with 10.2 for Mixing for awhile. R11 solved that issue though.

[mcatalao]

Are you also saying that there is a Big performance upgrade from 10.3 to 11 or am I just wishing?

Nah, the performance bump was on 10.3/10.4. 10.3 introduced the buffer rendering mode and 10.4 solved a bunch of issues with different vsts thst started crashing. 10.4 is pretty stable, but apart from some vsts r11.x and 10.4 have the same performance between them. I had then project I did to test 10.4 and ran it again with the same results.

[Le Boeuf]

Are you also saying that there is a Big performance upgrade from 10.3 to 11 or am I just wishing?

Nah, the performance bump was on 10.3/10.4. 10.3 introduced the buffer rendering mode and 10.4 solved a bunch of issues with different vsts thst started crashing. 10.4 is pretty stable, but apart from some vsts r11.x and 10.4 have the same performance between them. I had then project I did to test 10.4 and ran it again with the same results.

I see, thanks for the info, ill save the money for 11.5 then

[raymondh]

Yes the performance improvement was really welcomed!

I also like that you can 'turn off' VSTs to save DSP when you've bounced in place. Not the same as track freeze but still makes a huge difference.