WARNING! Long reply! tl;dr for @Grahf !
Well, for testing... and since you brought up the supposed issues, I/we thought you might be interested in this kind of test...
Alright, this could either be due to it simply being present on the original disc, or whatever...
If it where reading-issues it would replay the sound-sample whilst it is loading, due the buffer became so empty that it solely has one sample saved...
On another note: This short 'pause in movement' appears in a quite regular time-count of approximately ~1s... It is just more noticeable in some scenes with alot of 'one-sided movement' (left-, right-, up-, down-movement of the whole picture)!
Which is EXACTLY, what I just wrote there... WinHIIP INTENTIONALLY fills the gaps between partitions and I know, it MIGHT happen on a formatted/'fresh' HDD as well (that it produces a fragmented install).
But you should take a look at the answer you were quoting and to what I was replying... So your reply is partially out of context because I was replying that it intentionally does fill gaps, thus this is not an issue but a wanted behavior by WinHIIP (or it's creator)! There could be an issue about the 'fresh HDD'-Thingy tho'.
I agree to the laser and it's ability to read a DVD9 is a possible reason...
I do not agree, that this is the likely reason, tho'.
Well, I did not say it must be the reason but it seems likely to me.
It still could be something else, tho'!
I will try it from disc and SSD (possibly HDD as well) if/when I get it today!
The Bitrate is not soooo high... I think it was ~6MBit/s (~0.8MB/s) [MPEG2 permits up to 10.08MBit/s] on a 'CBR' instead of 'VBR'!
So you just claim unverified things (again)...
This is not really 'stuttering', which would imply that it has some slowdowns in some short scenes along multiple frames!
This is rather a kind of 'stop for a short moment and continue playing' and it is continuously happening after a certain amount of time!
It's similar, how simple converters convert between 44100Hz and 48000Hz...
I even ought to remember, that I did notice these subtile 'stutters', awhile ago myself (years ago)!
AFAIR, 'back then' I also thought it could be my laser, but later on I tried some DVD+R DL and it could read it without any problem so I doubt that the laser was at fault.
Small side-note: AFAIR you can keep the quality of the clips (in GT4) pretty much the same and still rip it to a DVD5, via ripping the Videos, but using a VBR instead of CBR!

The rips you will find on the net still use a CBR and are not really well converted... Their quality sucks compared to the original disc!
I agree! It is UNLIKELY a kind of fps-'drop' (at least the kind he thinks of)...
Where DAFUQ do you know that?
Even if the game had slowdown/fps-drops occasionally, this would NOT imply that these drops you can see in the videos are actually from the game-engine!
That's just another non-verifyed claim from you...
So does that prove or even imply, that the video-stutters are actually recorded framedrops? NO, it doesn't!
If that were the case:
- Why doesn't it drop the frames in the same way, if you play these actual tracks and are at the same point on the track and do the same like in the video?
- Why does it 'stop&play' continuously?
- Why does it do that everytime within a specific time-frame (which voids the fps-drop-idea entirely)?
- ...and why the freakin heck is GT4 able to render in a higher internal resolution without dropping more frames (i.e. in 1080i it used 640*540 or so)?
That's probably due to the fact, that you were not explaining your case thoroughly enough...
Interesting to know!
To me it seems that it happens the whole time, but is just more visually observable on whole-screen-shifts/movements instead of the 'zooming'!
Indeed... Occasionally, but not continuously on a quite 'timed' basis!
Thank you for testing that!
Oh wow, you already debugged it!
Great! So why are we still talking about GT4 then? ^^
You have no right to tell him, what he should stop talking about...
If you don't agree with it 'fine', bit you have no right to force your view upon others...
That's (quite) bullying! Stop that!
Well, a true comparison between PAL on HDD and DVD9 or NTSC on HDD and DVD9 would be better and can ultimately void his UNVERIFIED CLAIM...
THX for testing that as well...
OH REALLY! WOULD YOU MIND PROVIDING
ANY PROOF FOR YOUR CLAIM?
Ooooh! You already know for certain, that your assumption is true! That's great! Less work for others!
Would you mind sharing, what's so 'special' about that curve?
It is not?!
Then you should PROPERLY explain to us, what you are talking about...
Claims, claims, claims... and 0 proof... I thought you did learn a bit...
Oh really? ^^
Haha...
This does not mean, that the stutters where not there to begin with! Maybe you just didn't notice, because you either did not pay attention or because it simply is not noticeable ON YOUR SETUP...
Weren't YOU the one/person, who told HIM to not talk in absolutes?
(It's a rhetorical question... The answer is obviously: 'Yes, you are/were...')
Oh, sure... The pulldown isn't, but the freakin video-playback... So a PS2 is not even capable anymore, to play back an MPEG2-Stream? Lol! You ought to be kidding, right? I'm not saying it is not possible (maybe OPL affects it), but then it should definitely not happen on the original disc!
That doesn't
PROVE jacksh*t, but solely COULD
INDICATE a relation... but it does not
need to be related...!
Hadn't you just stated, that you were NOT talking about these 'recorded frame-drops' but 'the video playback'?
Aaah... I knew I did read that part right!
So how is the issue in GT3 (you are experiencing), related to GT4? Is it related and how so?!
Aaah... There has to be... Another assumption based on the premise, that the issues in GT3 and GT4 are even related in any way! Do you have any indications and proof for that (claim), beside it being the predecessor in the series? You haven't provided any (evidence, etc.) for that (claim) as well...
Are they? Have you analyzed their Bitstream or have you demuxed (de-multiplexed) the videos or how do you come to that conclusion?
How do you even come to the conclusion, that the Bitrate is in any way related to any stutters based on the transfer-mode and that the transfer-mode is not just a trigger (on your setup) for the root of the cause/issue?!?
Interesting! It would be quite nice to know which games absolutely need a coldboot-start and which experience problems after an IGR as well (even if they normally work without a coldboot-start)!
They might not directly be related, but there COULD be a link!
IMO, it is recommended to turn it on (for various reasons), but playing around with it for tests is alright.
As you can see (now), he made a typo and did actually try to disable it.
Up to this point everything is correct! It could have an influence indeed!
That's an unverified claim...
It also could be related to init and or the transfer-mode being preset, etc.! There are various other possible reasons!
Do you have a modchip? Else you would have to swap, if you'd try it from disc... and that will not show the Logo...
That is NOT true!
Pressing reset or rebooting FMCB via Software, COULD leave FMCB also in another kind of 'state' but on FHDB it is a bit more likely to trigger something!
There seems to be a few games which also only work after a coldboot-start in FMCB or at least not after running the hacked OSDSYS!
You/We don't know it for sure!
There actually could be an issue!
I still suspect FMCB to have a little quirk possibly related to the system-initialization.
It might have a connection with the OSDSYS and the PS2 Logo somewhere!
Good to know!
Are there games which work through ESR via a coldboot-start, whilst not working when the OSDSYS started up ONCE? I suppose there could be, even tho' they should be very rare!
Some modchips (DMS3v2 for example, if I remember correctly) seem to allow that curiously!
That very well could be true, but does NOT mean that FHDB does not have additional issues/things, which could cause this!
THX! It was one of those games, where I was interested in.
When you do, you should also make these tests on the ifcaro-builds, because the DBs might either fix or break (probably rather the later, IF there is a difference) some games.
AFAIK it did have some small code-changes (i.e. some change from l.oliveira was copied to it in some old build, AFAIR) which could have an influence!
Yes, it should not IMO...
Does ANYONE here have a PS2 which is not hardmodded and additionally has a Swap-Magic-Disc, to start OPL through that (logo would be shown, right before SM starts)?!?
That might shed some light on that particular issue!
This would be interesting for the IGR-Tests as well!
I'd really like to see that! I ought to remember that her face was kind of corrupted with distorted square-blocks of wrong colors.
Could someone make a video of that glitch?
Unfortunately, I think some part of our software in the loading-chain could have a bug, which is merely triggered by some game's code. This does not exclude the possibility, that this possibly existing bug probably could also be triggered by bugous game-code.
Is it obvious for THIS GAME?
Do you already know that this is one of those games again, which can not run in every DMA-transfer-mode?
Matter of fact is, that it would be the best IF every game WOULD work with every DMA-Transfer-Mode...
Why? Does it need a 'cleaner EE Kernel' and for what?
If RRV does a lot IOP-Reboots, this could make it even incompatible with OPL, because it can not further communicate with a backend/parts it usually has hooked in the EE-Kernel (at least I think it has/does).
Hahahaaa...
I agree to all of that.
The scroll-wheel now has to be the scapegoat? Seriously?! LOL
You are right about these, but if it doesn't 'hurt' but rather improve the overall game play, I don't see any issue with applying it.
Whuuaaath? That's definitely not true! Proof? See
@Tupakaveli's post about using it in GT4!
If that (what you are claiming here) were the truth, it would neither be a mode nor 'off' by default...
Why do you think these Syscalls are hooked to the EE-Kernel in the first place and why are we not unloading it on/for default?
You are CLEARLY claiming something false here!
That is NOT, what this mode does and is about, or how it works!
I agree!
Interesting!
Indeed! There MIGHT be cases (other games) however, where that applies!
:thumbsup:
Actually Mode 2 seems to be the second-most needed mode, after IGR (Mode 6)!
Mode 3 is way less often needed!
Where do you base that claim on, that only a very few need it...?
You YOURSELF have said, that there is a difference between 'working' (game starts and possibly works throughout the game, but might have issues like sound or FMV-stutter) and 'working' (perfectly without any issues)...
Well, it is a 'mode' for a reason and also it is disabled by default, because the smaller amount of games would need it, compared to the current config (async reads being quite compatible to a lot/most games)...
Mode 2 'weird', if it where only needed on one device (for the same game on the same setup)? No, not so much... But I believe there might still be other issues, which possibly are being circumvented, via using Mode 2... I suppose we will see less games needing it in the future!
All true...
I like that answer!
Well, it has to be confirmed to be compatible 'in the end', by someone who actually plays through the whole game, via OPL. But for testing of a game with issues, that's a bit too much to be expected from testers...
Actually, even tho' you probably over-exaggerated your example, you are quite close to the truth!
Different places in the game MIGHT need different 'bandwidths'... and you are right, it would solely mask the root of the issue!
Huh? What do you mean? Which game and which transfer?
Agreed...
IMO not Mode 2 itself, but there could be less circumstances which cause games to be dependent on Mode 2 than on earlier builds!
Aaaahahahaa... :'-D
True...
YMMD!
Was the recent change 'drastic' in regards to structure?
Those kind of TRUE fixes, actually are way better than what you are suggesting in that regard!
Correct! It does NOT fix anything but solely LIMIT to work around the issue...
Good!
No it is PER DEFINITION a 'workaround' because it simply works around the issue and NOT fixes it, thus this is PER DEFINITION NOT a fix!
It doesn't mean jackshit how a workaround is implemented in conjunction with a mode...
It is still neither a fix, nor would it work on all setups!!!
I am not claiming that I know all of it's structure, but you certainly know too less about the internals of OPL, to make any qualified statement/suggestion regarding differing implementations... You are not the only one, but you are the only one which ever claims to be right and demands it/things to be implemented 'his way'.
Does that mean the recent changes will not have an impact on this or any other mode, just because the direct code-part of it was not changed? NO!
The past and present already proves you wrong!
Not weird at all, given that I assume mode 2 being affected by threading (even after the recent fix, which was not a threading-change) and SMB-Support is VERY heavy on the IOP, especially regarding it's amount of CPU-Cycles (but also IOP-RAM)! 'Block Device Manager' by @Maximus32 might eliminate this quite a bit.
'Inadequacy of ETH'?
Well... That could be a reason, along other possible issues like threading...
The rest of the post is factually correct.
EXACTLY!
If someone tracks down the revision this started to be the case, then we might find the issue... THIS is a good example of what could happen via differing timings or related things... (which doesn't mean that this is the case here)
Maybe it started with one of
@sp193's 'mega-commits'?!
Interesting! (...and when you mark it in bold... 'It does not'...

)
True! AFAIR, the HDD-mode's CPU-Load on the IOP is rather small ESPECIALLY compared to SMB, but also USB...
Comparing the devices:
- Most CPU-Load (also during transfers): 1. SMB; 2. USB; 3. HDD
- Highest bandwidth: 1. HDD; 2. SMB; 3. USB