Dolphin, the GameCube and Wii emulator - Forums

Quote:I'm not interested in doing any more testing just because you say I should. You'll just have to take my word for it, or not.

Imagine if science worked this way. "I can't produce any data, you'll just have to take my word for it".

Quote:What is stupid is to not find out for yourself. The idea that '"I read it somewhere so it must be true" is not a valid argument.

Nice to know. I guess I'll just throw out all of those scientific journals tomorrow. Because you know, data doesn't count unless I collected it myself.

(10-12-2012, 04:06 PM)NaturalViolence Wrote: [ -> ]Imagine if science worked this way. "I can't produce any data, you'll just have to take my word for it".

I think you're confusing me with someone who is obligated to do something. Obviously I have nothing to gain by saying what everyone already knows, that LLE is much slower than HLE.

Quote:Nice to know. I guess I'll just throw out all of those scientific journals tomorrow. Because you know, data doesn't count unless I collected it myself.

Obviously that doesn't apply to something that you can personally verify yourself, and that would be the smart thing to do before forming an opinion about it.

"Opinions" implies that I have no reliable data to base this on.

This is not politics, this is science. Define a question, make observations, create a hypothesis, test/collect data, analyze the data, draw conclusions from the data, publish the results, and retest if necessary. Every time we make one of these investigative threads we are following the scientific method whether you realize it or not. It is the only reliable way to form conclusions. You're not obligated to do anything but without any evidence whatsoever or specific complaints about the methods used for testing you have zero grounds for questioning the results. So either prove them wrong by contributing to the test or get out of the way.

I can confirm what Starscream said is true (um...Probably ) . LLE is significantly slower or not depending on your CPU
Last year , i tested LLE and HLE with SMG via desktop , i got 30FPS with LLE and 60FPS with HLE
I also saw the same result on FX-8150 CPU . Somewhere in this thread
However , my graphic card (Asus 9800GT) was broken and LLE backend was speedup in recent dolphin version . So i can't provide the correct result right now
I've just tested Xenoblade on G41 chip , i got 17FPS on both LLE and HLE (not a single FPS difference even with lowest resolution) because of GPU bottlenecking although the game used to run fullspeed with Asus 9800GT , 3XIR, no AA, 16xAF , HLE
Full Specs :
E5300 @ 3.6GHz
Onboard G41 chip
4GB of Ram - Corsair DDR3 1333MHz

Shonumi: Thanks for the invite, maybe I'll do some testing if the mood strikes me.

(10-12-2012, 04:12 PM)Starscream Wrote: [ -> ]Obviously I have nothing to gain by saying what everyone already knows, that LLE is much slower than HLE.

Based on the results in this thread,that statment is simply not true for most ( if not all ) wii games.

---

(10-12-2012, 04:03 PM)Shonumi Wrote: [ -> ]Using Wii games with streaming audio significantly reduces the gap between LLE and HLE audio compared to GC games that use sequenced audio

This is what most results show-LLE in wii games is simply not as demanding as in gc games.

(10-12-2012, 04:03 PM)Shonumi Wrote: [ -> ](I've not been able to reproduce some of rpglord's results where the percent difference was extremely low).

Thats because you havent tested games where difference is extremely low
In last story and monster hunter tri there is almost no difference-it varies from 0 to 1 fps difference.

(10-13-2012, 01:47 AM)rpglord Wrote: [ -> ]This is what most results show-LLE in wii games is simply not as demanding as in gc games.

Again, while this is interesting in itself, we're testing the gap between HLE and LLE. We've shown that the gap can be greatly shortened, but it still persists and the numbers still can be considered significant or at least relevant.

(10-13-2012, 01:47 AM)rpglord Wrote: [ -> ]Thats because you havent tested games where difference is extremely low
In last story and monster hunter tri there is almost no difference-it varies from 0 to 1 fps difference.

That's what "some" means, not "all". The ones I have tested are for SMG (will post results later for different levels).

EDIT - Out of curiosity, I decided to do some testing on later revisions. On Yoshi Star with 3.0, the difference between HLE and LLE (No VBeam for either) closed down to an insignificant amount (77 FPS vs 71 FPS, +8%). This dramatically increased when I used my normal 3.0-416 build (76 FPS vs 58 FPS, +28%). 3.0 truly is amazing to lower the gap like that. I know this benchmark is for 3.0, but once I get more games dumped, I think I'll do a write-up on how revisions affect the gap between HLE and LLE.

Shonumi,thank you for contributing,I am looking forward to your results

So, I'm trying to benchmark SMG2 with Dolphin 3.0-776 but I keep running into a few issues. I wanted the results to be as accurate as possible, so this meant creating "movies" in Dolphin (.dtm files). Almost every time I try to replay the movie, it complains about a fatal desync and complains about the wiimote (not at my computer atm, will post a screenshot later). When it does work the movie does the wrong moves at the wrong times (Mario will crouch inside of running, for example, or run backwards instead of forwards). Tried using a keyboard config, joystick config, and even iController on my iPod, but all of the above problems still occur. Anyone have much luck with making .dtm files?

Try the latest dolphin :3