27 réponses à ce sujet

[Drumrboy]

 So, the main problems have now been fixed (THANK GOD!) but with the side effects of some small stuff such as the Hammerhead parts of the Project Overlord DLC still having rubbish textures.  But thats understandable.

So now I ask the question.....

WHEN WILL THE AUDIO LAG BE FIXED!

Considering that ME2 has got amazing sounds and music, the audio lag just takes away from the experience. This is the only thing left to fix. But I can't put my hopes up and I should be grateful for 3 awesome patches already, but I doubt a 4th will come out.

What do you guys think?

Modifié par Drumrboy, 06 avril 2011 - 07:31 .

[xvynx]

Audio sync is indeed the biggest problem so far but after the most recent patch the gensis comic decisions are not carrying over correctly and thats a damn important point in the me2 saga. So ya, bump.

[titan_warrior]

First of all we are in the 1,02 version. For me, this version is unacceptably bad for a 2011 game, even for a ps3 port. Both the "Insanity difficulty trophy fix" and the "New Game Plus fix" (which are the main fixes of the second patch) aren't the type of fixes that improve the mass effect experience. I was hoping for fixes concerning framerate issues, loading times, visual bugs, audio problems, collision detection issues and game stability. Both the framerate inconsistencies and the loading times worsen after 1,01 version. The collision detection issues are also important ("flying Shepard issue", "Hammerhead going through ground"). The audio issues are of course the biggest problem, as they can distract us from the story. The pixelated ground from overlord dlc might not be so important, but it is so horrible and unpleasant!!!! As for the game stability, it has been improved with patch 1 but I was expecting improvements from patch 2 too. I understand that these issues are difficult to fix or even improve (as they are ingame issues), but I hope at least Bioware will try. These are issues that need time and effort, but can improve the whole mass effect experience.

[RaynarBrusk]

I'd say the audio problems are by far the biggest problem short of anything that actually breaks the game. After all the hype, I was actually disappointed by Overlord, I expected so much worse from the complaining. For what it is, it's about as good as it needs to be.

[johnny vee]

I don't find the audio to be much of a hassle when you play with subtitles on.

[Dragonfan8574]

If i could only have one more fix in a patch it would be the audio issues. This dosen't count since it occurred from the new patch but the new genesis glitch should be fixed to and hopefully soon. If i could have two problems fixed it would be audio and the overlord texture issues. Those two by far are the biggest next to the new genesis one.

[GFX1989]

I think the fixes should go in this order.

genesis choices > dialogues getting cut off > sound desync during cutscenes > overlord textures

[Zaylem]

1. Genesis patch should be fix first since we want to carry that to ME3.
2. Dialogues and sound effect should be next because most of us play this for the well written story.
3. I could live with out the overlord texture being fix

and tnx bioware for the patch

[TheBisk]

Definitely fix the genesis choices issue soonest please.

[titan_warrior]

Two days ago I was certain that the second patch would be the last one but after the genesis glitch (created from the current patch) I've changed my opinion. I am certain now that a new patch fixing the genesis issue will indeed be available, as the genesis bug is easy to fix. Don't hope for audio fixes though as these fixes need changes to the whole code, which means a lot of time, effort and frustration. With Mass effect 3 being scheduled for a simultaneous release in December, an audio fixing is impossible. The schedule is too demanding for difficult bugs like this unfortunately. I don't expect much anymore.

[Drumrboy]

titan_warrior wrote...

Two days ago I was certain that the second patch would be the last one but after the genesis glitch (created from the current patch) I've changed my opinion. I am certain now that a new patch fixing the genesis issue will indeed be available, as the genesis bug is easy to fix. Don't hope for audio fixes though as these fixes need changes to the whole code, which means a lot of time, effort and frustration. With Mass effect 3 being scheduled for a simultaneous release in December, an audio fixing is impossible. The schedule is too demanding for difficult bugs like this unfortunately. I don't expect much anymore.

I see what your saying, but the patches before hand took so long since Dragon Age 2 and Mass Effect 3 were in the making. But now BioWare a more focused on ME3 I think that it should be easier, though I'm not putting my hopes up for the audio fix but one can dream :innocent:

[Drumrboy]

Drumrboy wrote...

titan_warrior wrote...

Two days ago I was certain that the second patch would be the last one but after the genesis glitch (created from the current patch) I've changed my opinion. I am certain now that a new patch fixing the genesis issue will indeed be available, as the genesis bug is easy to fix. Don't hope for audio fixes though as these fixes need changes to the whole code, which means a lot of time, effort and frustration. With Mass effect 3 being scheduled for a simultaneous release in December, an audio fixing is impossible. The schedule is too demanding for difficult bugs like this unfortunately. I don't expect much anymore.

I see what your saying, but the patches before hand took so long since Dragon Age 2 and Mass Effect 3 were in the making. But now BioWare are more focused on ME3 I think that it should be easier, though I'm not putting my hopes up for the audio fix but one can dream
:innocent:

[xvynx]

titan_warrior wrote...

Two days ago I was certain that the second patch would be the last one but after the genesis glitch (created from the current patch) I've changed my opinion. I am certain now that a new patch fixing the genesis issue will indeed be available, as the genesis bug is easy to fix. Don't hope for audio fixes though as these fixes need changes to the whole code, which means a lot of time, effort and frustration. With Mass effect 3 being scheduled for a simultaneous release in December, an audio fixing is impossible. The schedule is too demanding for difficult bugs like this unfortunately. I don't expect much anymore.

how exactly do you know genesis is easy to fix and audio is hard, i know my bit of programming so feel free to get techy

Modifié par xvynx, 07 avril 2011 - 06:20 .

[titan_warrior]

It’s really difficult for me to analyze my opinion but I will try. While in theory the Xbox360 has only half the pure hardware performance of the PS3, achieving the theoretical maximum is much harder with the PS3 architecture. Therefore many games will actually perform at parity, and for some games or aspects of games, worse. Games will also take longer to develop or leave performance on the table, and most likely have more bugs. The additional development costs, TTM delays, and lack of a performance edge will significantly impact the PS3. First of all lets talk about the ps3 architecture and how this affects the quality level of ports. I won’t use technical terms and I’ll try to speak as simple as I can. The ps3 has let’s say a main processor and several other microprocessors. This is called Cell (microprocessor architecture) which was developed from Sony and used in both ps2 and ps3, making these machines a real pain in the ass for the developers and I will explain this later. On the other hand, both xbox and pc have a cpu and a gpu connected via a motherboard. We are simply talking about a completely different philosophy here. Also pc and xbox have a decent amount of RAM, with which the stored data can be accessed in any order in a constant time. Now let’s get to the point…. Let’s say that a game is being developed for a multiplatform release, but uses ps3 as the leading machine. The ps3 has many microprocessors and each one can be assigned for a specific job. Being the leading machine, each of its processors will be used from the developers so the game will be running smoothly without bugs. The porting procedure to pc and xbox won’t be difficult in this situation, as it is easy if you have lets say seven separate peaces of code (assigned to seven microprocessors) to merge them into two (one for the cpu and one for the gpu). Now lets say that we have the opposite situation, were lets say Bioware developed a game (mass effect 2) with the xbox 360 as the leading platform and later tries to port this game on the ps3. This is a very different process and a very difficult one, because while merging different parts of code is easy, separating a code into many peaces and assigning them to various microprocessors is not only tough but also really buggy. This means that many ps3 ports don’t use all the ps3 microprocessors, as it is difficult to assign for example the whole xbox’s GPU operation to lets say three-four ps3 microprocessors. That’s why the ps3 sometimes struggles to run ports, as some of its power isn’t used at all. Also the process of breaking the code into smaller pieces creates bugs. So what is the solution for the devs? They must either use ps3 as the leading platform, or develop the ps3 port for a longer time and supporting later that port. The ps3 porting process needs a lot of time and effort in order to be both effective (more microprocessor power used) and bug free.
Now that we analyzed these lets go to your question. There can be two different types of bugs and glitches, those which are the result of performance issues and those that are result of bad coding and need debugging. When you apply a patch (which is a new code inside an old code), the resulting code might not even be able to work properly. Even a comma, a parenthesis or anything can cause chaos. These types of bugs are result of bad coding and with the debugging process the devs can find both where in the code (for example in line 2003 of the code) there is an error and what happens when that error occurs (for example does it loop, does it ignore a process and continues to the next one, does it crash and more). In this way they fix these bugs, searching for errors in the code, detecting the specific coding dysfunctions and correcting them. Examples of bugs that need debugging are the NG+Glitch and the Genesis bug and caused either due to hasty porting or due to hasty debugging. On the other hand, the corrupted data bug, most of the crashes, the framerate issues and the ps3 incapability playing simultaneously the voices and the effects during conversations (audio bugs) are mainly caused due to the non sufficient usage of all the microprocessing power of the ps3. In order these bugs to be dealt efficiently, some processes must be assigned to different processors, which means of course a complete rewriting of the code! If many operations are assigned to a single microprocessor, the ps3 performance will be disappointing. The only solution is to divide them and use more ps3 power, which mean that the porting procedure will take longer to complete and might also need some extra debugging afterwards for dealing with the other types of bugs that will arise. The debugging process is hard and difficult, but it cannot be compared to the assignment of different operations to different ps3 microprocessors, which is the real pain in the ass for the devs. That’s why the audio issues are much more difficult from the Genesis bug. I hope I helped.

Modifié par titan_warrior, 07 avril 2011 - 08:01 .

[xvynx]

man reading that on my modbile was HARD, bioware needs to drop the white font on black background.

Neways i see your point, some logic behind it indeed.

[SirBrass]

Titan, I'd believe that about the audio issues if itd been there since the beginning on the PS3. It wasn't. It wasn't until after the next to latest patch that it started happening (at least for me). The Genesis glitch didn't crop up until yesterday's patch.

If the issue wasn't there from the beginning, then it most likely is NOT an issue with inefficient splitting up of the code onto different SPUs.

We need to go back to when the issues started happening. For me at least, the only bug I had starting out was the random crashes (which was bad... really bad... on the same level as FO: New Vegas bad, and not good for the hardware to have to be physically reset at least a half dozen times a day during an 8 hour weekend gaming marathon session). No NG+ issue (never saw that, honestly... maybe I'm lucky). The audio issue happened (slowly) after the first patch and is difficult to reproduce for individual instances but happens frequently overall.

Okay, so first major breaking issue that didn't occur until the first patch. This could be an issue with how the new fixes integrate with the old code (new wine in old wineskins and all that plus a dose of Murphy is quite possible, and the worst actual cause for a systems engineer/QA debugging engineer trying to specifically diagnose the issue and send it down to software with specific instructions on a fix), OR (what should first be examined thoroughly until ruled out by logic) there is a bug in the new code. This is easier than examining ALL of the code itself. Essentially such a diagnosis comes down to doing a compare of the pre-patch code to the post-patch code. Examine the differences and investigate those sections for a bug or innately tame code which can cause issues. Maybe someone wasn't as careful with a new section of code and it is eating up too much memory and the system has to intervene to free up memory at a critical point, and the audio somehow gets decoupled that way... I don't know. But it is obvious that the audio start/stop (at least the stop) is bound to the end of the speaker's mouth-movement time hack (that is, when the person's mouth stops moving for that section of speech, the audio is stopped in the code, regardless of whether or not it has reached its end). I'm tempted to posit a solution, but without actually knowing WHY the issue occurs, a solution offered is not worth much and could just jam up valuable processor resources when a more elegant solution is viable.

Now, the genesis bug. This should be even easier to debug (though trust me... that doesn't exactly mean that it'll be easy to FIX). It crops up (at least for me) after the most recent patch. Its symptoms are that backstory choices are FIXED regardless of selection. The issue is that things like codex entries are still accurate (at least from what I've seen), but in-game dialogue is not.

Lets take my specific issue as an example: I chose to save the council in Genesis (I figure that in ME3, a grateful Turian Hierarchy will be valuable what with all their dreadnoughts available, whereas a PO'd galaxy who are mad at humans for the political coup will be less likely to aid Humanity figuring our demise was fitting justice for us screwing them over in the battle for the citadel). However, ALL dialogue except for speaking to Anderson at the Presidium is that which is used when you chose to NOT save the council. The reporter, the turian at the store, captain bailey, etc. Though the scene with anderson is still what it would be for the choice made during Genesis... until you encounter Udina.

This still has all the markings of default values being written instead of the values input from the interface, and that would be the first place I'd look if I were debugging the post-patch source code, and examining it against the original code and the post-first-patch code to look for discrepancies which could cause this.

I can't say anything less vague than this, as 1) I'm not a bioware engineer and thus don't have access to the code to actually do the analysis. and 2) b/c of #1, I don't have access to the software suite which would allow me to test my hypotheses.

Just a brainstorm from a guy who is used to having to debug issues where one has to look for errors in requirements, hardware/software integration, and the software itself. Figuring out where to start is the most important part. And as far as that goes, the more recent the issue first starts appearing, the less code one has to dig through to generally find the issue, or at least find the problem area.

I hope that if the bioware guys read this they're thinking "duh! we're already well along that path!" Means less time till a fix comes out.

[titan_warrior]

First of all let’s talk about the Genesis glitch. The digital comic code is in the very beginning of the whole code. Now although a code written on a C++ (object oriented language) has several different parts (digital objects), where each of them is being used in the main code, let’s assume that there is only the main code. The digital comic’s choices are reflected on a very early part of the code. If you don’t download the digital comic, the choices of mass effect 1 are fixed (I assume it’s “Kaiden dies”). Now let’s assume that in line 100 of the code we have the following code: if object(digital comic)=false then “blue” else {if [object(Kaiden) selected] then “blue” else “red”}. Now later in the code (when the game searches the sacrifice choice), the ps3 will search feedback from line 100 of the code. Initially it will see if object(digital comic) is true or false, which means if you have downloaded the digital comic. If not, it will use the fixed answer blue which means Kaiden sacrificed. If the digital comic is indeed downloaded, the ps3 will read the next line which is if [object(Kaiden) selected] then “blue” else “red”. If “blue” kaiden died, if red Ash died. Something happened with patch 2 and ps3 assumes that object(digital comic)=false which means that digital comic isn’t downloaded. Whatever happens here, all Bioware must do is debugging. They know where to look so it’s relatively easy.
Now let’s talk about the audio issues. I think that the audio sync was initially problematic and that the audio cuts where introduced after patch 1, but I am NOT sure. I am sure though that after patch 1, the game play WASN’T as smooth and frame rate inconsistencies where introduced. This means that in order to secure stability (as crashes created the data corruption bug), the code became heavier and couldn’t run efficiently from the same microprocessor. This means that even if initially the ps3 could smoothly run the code, after the necessary changes for securing stability ps3 couldn’t run the game without using more of it’s power (more microprocessors), as the code is now heavier. This also means that a rewriting of the code is essential in order to better exploit ps3 architecture. Initially the full usage of ps3 power wasn’t a necessity (if audio issues weren't present before patch 1) but things have changed since patch 1. Unfortunately splitting the code into more processors needs time and can create chaos in the code’s logic if not done carefully. Also a debugging procedure must follow and testing is needed afterwords. In other words we are screwed. Of course this is my opinion, I hope I am wrong.

Modifié par titan_warrior, 08 avril 2011 - 01:30 .

[Matt VT Schlo]

Please fix the load times on the Normandy 2. shouldnt take (what feels like) 20 minutes to go from level to level

[InfamousDarkDragon]

I think the Genesis comic choices and the cut scenes audio lag are the most critical. I actually thought that 1.02 would solve my problems, not create new ones.


- Fix audio lag in some cut scenes (Freedoms Progress when the Heavy Mech is shooting at the Quarians)



Time in @ 3:13


- Fix all audio cut offs when characters are talking.

That's it for me!

Modifié par InfamousDarkDragon, 08 avril 2011 - 07:53 .

[WhiteWi]

Agreed.

InfamousDarkDragon wrote...

I think the Genesis comic choices and the cut scenes audio lag are the most critical. I actually thought that 1.02 would solve my problems, not create new ones.


- Fix audio lag in some cut scenes (Freedoms Progress when the Heavy Mech is shooting at the Quarians)



Time in @ 3:13


- Fix all audio cut offs when characters are talking.

That's it for me!

[titan_warrior]

When I bought the game I thought that it was a game that its porting process took almost one year, as it was released a year ago for the other platforms. It seems now that this “one year in development” thing is a complete joke. Bioware was simultaneously developing dragon age 2, mass effect 3, various DLCs and porting ps3 mass effect 2. This port was hasty and without any care. Is this the right way to gain new fans Bioware? The only thing that saves you is that the game is fantastic, completely from another planet. This doesn’t mean that the specific port is good. A game can be both fantastic and a bad port. The whole situation of “how many patches do we need”, “what glitches must we select from all these bloody bugs to be fixed”, “what new bugs did the last patch created”, this situation is completely humiliating and frustrating. You only cared for fixing the “save corruption bug” as it was leaked to the press (and was also a game breaking issue) and then continued with the development of the arrival dlc and mass effect 3. When the ps3 fans started to protest for the mess you’ve created, you supplied a quick patch (developed during the week that followed the arrival release) so we shut up. I understand that the ps3 version of mass effect 2 didn’t sell well, but this isn’t a way to treat your fans.

[titan_warrior]

Something must be done with the “save corruption data” bug, cause it is still there, “alive and kicking”. As I’ve wrote previously, this bug is a performance bug. There is a reason why Bioware advises as not to play the game for a long period. When many operations are allocated to specific ps3 microprocessors, the ps3 struggles to run the game so ps3 becomes overheated and therefore crashes become more possible. In other words the corruption is linked to the crashes and the crashes are linked to ps3 performance, which is linked to the fact that some microprocessors remain inactive. Therefore ps3 struggles to run a game not because it hasn’t the power but because Bioware doesn’t exploiting ps3 full power, which results performance bugs such as crashes, corruption and audio issues. Therefore a rewriting of the whole code is essential in order to take advantage of ps3 architecture. Bugs that are caused from “bad coding” such as the NG+ and Genesis can easily be dealt from the devs via the debugging process, but performance bugs need a complete rewriting of the code, which means time, effort and frustration and can cause new bugs to arise that need debugging. Performance bugs cannot be fixed via a quick patch like the first one that supposed to fix the “save corruption bug”. The problem is that voice issues, save corruption and crashes are three issues that although linked to performance (so are difficult to be dealt) must be fixed as they are too important. I believe though that only the Genesis glitch will be fixed, as it is a bug that only needs debugging.

[Alakratt.0]

After this patch notice, I would call them, "quick freezes" when a lot of enemies appear. This didn't happen after the previous patch.

[MarkakaJin]

Would be great for Bioware fix the audio lag. That's the main frustration about this game.

[Valkyre4]

The audio lag is really unacceptable in such a game that is 50% played through cutscenes and dialogues. I mean what the hell? I couldnt care less if the audio is not synced during cutscenes in a game like Call of Duty, but Mass Effect? Half the frigging game is about cutscenes!

2 patches already and this simple thing is not yet fixed. And no, I wont believe that it is so hard to fix audio sync...seriously wth?