Test Design: Aleksey Lukin...
Yes, I know that... Actually, I've even able to improve the current algorithm a little (only +3 dB of SNR), but it shows better results (no spikes all over the spectrum) on the 1 kHz sine test. You may check this out in beta v1.2.9.
I've already commented that SRC test at the Discussions forum.
In generally this test has some flaws and questionable things in my opinion.
Even if flawed, it still shows up how good R8Brain Pro really is.
The only one that looks to be better is the Weiss Saracon, and that is some seriouds money - plus, all is still below -150dB with R8B, so inaudible in a 24 bit system that has a DR of 144dB in a perfect DAC, which as we all know does not exist.
I would be interested to know what the flaws are though.....
Multichannel Audio Specialists
DVD-A, DVD-V Authoring
Mixing & Mastering to most formats
Please see this topic: SRC Shootout
It has been approved that their benchmark is flawed. This flaw has been fixed already.
Moreover, as I have found out r8brain free performs quite well in 1kHz sinewave test (equally good to Saracon) and if I could improve its antialiasing filter (make it steeper and fine-tune the transition band), it would compete with other converters which are not 'least common denominator' ones.
It will be interesting to know whether or not Saracon is least common denominator type as well.
BTW, the latest v1.3.1 features speed improvements which made ultra-steep mode even faster than non-ultra-steep mode of previous versions. Non-ultra-steep mode of this version 1.3.1 is twice as fast. And I have plans to implement multi-threaded conversion which should rise speeds by a factor of 1.5-2.0. That should make the conversion even faster than r8brain free method (which is meant to be pretty fast) while offering an 'ideal' quality.
Looking very much forward to these improvements!
This topic was last updated 180 days ago, and thus it was archived. Replying is disabled for this topic.