Input Lag: A Scientific Experiment (Epson 8350, 3010, 5010 & more) - Page 5

Sure enough. Now I just need 12 of these end-to-end and I'm on my way!

My guess is that the Amazon guy is telling the truth, BUT the reason he sees a speed difference is not in the length of the cable, but rather that the shorter cable just happens to be of better quality.

So, my theory is that the long cable is cheap and crap, and that the communication between PS3 and Projector is having some errors (it is a digital connection, not analog, so there is a handshaking protocol and a data correction check etc going on every few packets of data).

Now, since the 'lag' seems to be a common issue for most everyone with this PJ, perhaps the Epson is at fault, and just doesn't have enough tolerance in the communication for some reason, so the longer/cheaper cables cause it to balk?

I once had issues with a Sanyo projector on a 25 ft HDMI run. It actually caused the screen to flicker, so I upgraded to a larger gauge cable, same length, and never failed again. Who knows, maybe 'failure' can come in the form of lag because of numerous packet retries?

My guess is that he had some super crappy cable that gave a bad picture and now he's replaced with a shorter cable that works better... but has the lag that the rest of us are still complaining about. Because I don't think the problem is in the cable.

EDIT: He posted a youtube video: http://www.youtube.com/watch?v=4tONKWbfnHA Maybe he is onto something. Some numbers would be nice. Someone who can actually test this should run some numbers.

I'm pretty sure HDMI is a "transmit and forget" protocol. Meaning the receiver either gets the data from your Xbox/PS3/Bluray or it doesn't. If it doesn't get the data, you get color sparkles or worse.

I don't believe this "cable solution" has merit at all. But, I'll be happy to admit I'm wrong if the data shows it...

... Altan

Well I am no HDMI expert, but I do know it is a 2-way protocol, and that there is an initial handshake at the very least. Everything else I said in my post was slightly educated speculation.

I agree it's a two way protocol, but it's not a two way protocol with regard to sending the video data. The control data is 2-way, but not the video data. To be very clear, it's not like the projector says "hey! I missed that video data, please resend it.". If it's lost, it's gone.

... Altan

Interesting. Too bad I don't have this projector to test myself now. I have a 1 foot HDMI cable however (Slams head on desk). A confirmation from about the firmware / fix from Epson would be nice. Thank you for sharing.

Lag on the epson projectors really isn't debatable at this point. On the HDMI cable thing-
Fact 1: numerous people have tested projectors with very long hdmi cables having minimal/immeasurable lag.
Fact 2: numerous people have tested the lag in the epson 3010 projector, with various configurations of HDMI cable length, to have high lag (I myself have tested it at over 60ms compared to a slow LCD monitor, then hooked the exact same cable to a pro350w and got 0 lag)
Fact 3: This lag has been confirmed in a professional, published review (art of gaming)
Fact 4: Epson themselves have confirmed the lag exists, and that it was designed into their specs (this is confirmed in their response to several customer support requests, including both mine and their response to the above review)
Fact 5: it's even written into the extended EDID information in the projector (extended EDID states it at 125ms)
Fact 6: HDMI cable quality is vitally important to image quality - in that if it fails, the image (or portions of it) will simply not be there. The HDMI communications process does not allow for re-sending of data in an active video stream. Therefore, there is no possible explanation for lag relating to HDMI.

With all that in consideration, the idea that a shorter/longer hdmi would fix the lag on a 3010 is provably false many times over.


I could test this myself, but it would be a pain since I'd have to find a way to put my computer very near to my ceiling-mounted projector box. I could also climb onto my roof and throw off a bowling ball to see if it would begin moving downward as described by the theory of gravity. Neither tests are worth my time as the outcome is already known.

Yikes, what's happened to this thread? This is AVS Forum, the S stands for science, people!

I love repeating what other people wrote, so I'll do it again: Cabling won't make a damn difference for input lag. That person is delusional and in denial. Those projectors have bad lag, end of story. Copper is copper, the electricity gets there or it doesn't. It's not like TCP/IP, there is no resending, no way to slow it down.

Defiancecp and tommyv2,
I agree(d) with both of you. Did you watch the video? Something is going on.

Get the results first, explain the science later!

Something I'd try if I had the 3010 still and a short cable... I'd use a long cable to a splitter or switch and then use the short cable to the 3010 and see if that works.

I agree, it shouldn't work... but the video has me convinced that the guy's not completely brain dead or confusing input lag for something else.

Best-case scenario is that the signal would travel through the HDMI cables at the speed of light. Light travels one foot in about one nanosecond. Therefore, it IS accurate to say that a 1 foot cable delivers a signal to a projector more quickly than a 6 foot cable, or a 35 foot cable, etc. However, we're measuring lag in terms of milliseconds, not nanoseconds. So if a given projector has a lag of 80 milliseconds with a 35 foot cable, that is 80,000,000 nanoseconds.

So if you go from a 35 foot cable with a lag time of 80,000,000 nanoseconds, you can expect the lag time from a 1 foot cable to drop all the way down to 79,999,965 nanoseconds, which is equivalent to going from 80 milliseconds to 79.999965 milliseconds. I am not an expert at human reaction times, but I tend to doubt that such a drop in lag time would be noticeable to any human.

I'm not saying that there isn't a difference from using the different cables - there could be some explanation for that.

I watched part of the video until the guy's random throwing-around of the camera started to give me a headache

I'll give it a full watch this afternoon.

Still, nothing in any youtube video is going to change the above facts. If he's showing two different lag results, it's not because of the length of the hdmi, and it can't change the fact that the projector was designed and tested in a number of settings to have significant lag.


Incidentally, my first round of testing was done with a 6-footer if I remember correctly (back with I just had the projector on the floor)

No, but it might change the way we interpret them.

But maybe that's just the hope talking.

Well I'm going add my 2 cents! I saw the 3010 on sale at BB(black friday) my tongue was hanging out and was about to pull the trigger when finding the problematic Lag thread that made gamers swoon around the world, I gave up,to bad maybe next year. Any how my beautiful wife surprised me on Christmas. Ignorance is bliss ---I mounted it on a step ladder hooked up Time Warner cable box(their hdmi cable) pany 210 bd(insignia 13.0 Gbps
High speed hdmi) and my Wii (rocket fish component) and when I play Mario Kart IT RESPONDS FASTER than my Sony 34 CRT and way way faster than my 52" 750 Samsung LCD! So what does this mean, low res beats hd? Hdmi - vs- component? Component can handle 1080 p no? And why would faster (480 Hz) refresh rate have more lag?? I am enjoying the 3010 and it feels faster,not slower!?

Look, I'm happy for all the people that either simply aren't bothered by the lag, or that don't believe it is there, etc. That's awesome. And I've said from the beginning, some people won't notice or be bothered by it.

But every single OBJECTIVE test done to date has shown it there.

I think it's important that people hear from those that can't tell it's there, just so they know it's not a black & white issue, and gaming is NOT completely incompatible with this projector.

But on the other hand, I think it is harmful to further the idea that there's some magic bullet that can get rid of lag that epson DESIGNED into it, and they themselves confirm cannot be turned off, and every objective test EVER DONE on it confirms to be there.

When someone does an objective test that shows it gone or nearly so, man I'm going to be jumping up & down with excitement. Until then, it's there, and some people thankfully won't be bothered by it a single bit. Objective test > subjective observation.


Again, I'm not pressing this point to argue with anyone or anything like that - I just don't want someone thinking they can buy this projector and somehow get rid of the lag. I do want people to understand the lag may not be a problem and buy this projector.

Amen. When it comes to people that don't notice or care about input lag - I don't know if I should feel bad for them or if I feel jealous that their ignorance is bliss! I base 100% of my display component purchases based in input lag as the #1 most important thing, and it's something I noticed before the Internet starting talking about it.

Hell, back in 2004, I worked at a day trading company - some of the guys got LCD monitors, and they swore that there was a split-second lag of their real-time feeds and stats. In a giant room with 100 monitors facing you, you could see the CRT monitors change picture first! You can't be in denial when the science presents itself to you with your very own eyes.

Panel refresh rate has nothing to do with input lag. The lag's happened well before video makes it to the panel itself - that's the entire problem. If your 3010 feels faster than your old display, then I'd like to see you game on a Sony HW10 and have your socks blown off!

These tests are objective but take these subjective test environments:

1. Is the CRT and Projector being fed from the same PC video card or separate?
If separate, then there could be an inherent delay between th two outputs.
To test this, you could swap the two outputs.
2. Even if they are using the same PC video card, is there an inherent delay between the 2 outputs?
Meaning if they switched connections between the projector and CRT monitor, does the results change significantly.
3. Are they using HDMI direct to the projector and VGA to the CRT or using adapters?
Do these two type of outputs may have inherent delays.


*I believe there is a delay but the measurements may not be exact.
That is why some can tolerate them while others can't.
It may depend on the device(Xbox, PS3, PC) or specific game being used that causes more delay than others.

Maybe not quote "nothing"...

It does make sense that if you have to process the picture from 24hz, 30hz or 60hz to 480hz that that would account for some of the delay before the picture is sent to the screen.

Especially when you take CFI into account (which the 3010 doesn't do). It has to receive frame A and B before it can display frames A, A1, A2, A3, A4, and frame C before frames B, B1, B2 etc...

Defiance, did you look at the rest of that youtube video yet?

I'll still go with "nothing"

There are 60hz CRTs that have zero lag, and 60hz LCDs with lots of lag - refresh rate itself is not an issue. Even the marketing "600hz!!!! omg" plasmas can have lots of input lag. The refresh rate is only tied to the rest of the technology and what they want to do with it - nothing has shown that any "fast refresh" products have improved lag - the opposite, in fact, but only because of the processing required. The high end HP ZR30w monitor is strictly a 60hz panel and has 10ms input lag, and that exact panel is used in the Dell that I have which does 25ms lag. It's all about the brains that power the brawn!

I'd like to say that I could not agree more. I followed after defiancecp and got the 3010, (also got a 9500), measured the lag under various conditions, and played various games on it. Input lag is right there on the 3010. I, too found it acceptable on 3D PS3 games like Uncharted 3 and Batman, but found it intolerable on a simple 2D pinball game. It will depend on the gamer and the game whether the 3010's input lag is acceptable.

While there might be somethings one can do with cables to make the lag worse, there is no way to make the HDMI input lag better than, say 80 ms. on the 3010 short of a major firmware tweak, which we should not hold our breath for.

The 3010 remains my projector of choice for value.

with all due respect to difiancecp for his objective take on it, i am yet to see the lag that is being talked about. I play FPS games likes Uncharted 3, COD-BO, Ressistance 3, KZ3 and other games like GOW3, Avatar 3d, super stardust, motorstorm and never once felt the lag in any shape or form

the only time i felt lag was when playing the Infamous 2 demo long back where Col's movement was jerky and the aiming was a bit of wobble as i tried to compensate for what you might call lag

so yesterday i got the famous 1 ft HDMI cable and i felt it was faster than my 50ft cable. I tried the Infanous 2 again and i saw absolutely no lag and the game was smooth. I am not sure I am just experiencing a placebo effect though

I dont have a HTPC so I am not sure how to do an objective test. I do have a Dual HDMI Samsung 670 bluray player and if anyone can suggest a good way to measure lag let me know and i will try.

All of you, please do this: Download the Mega Man 9 or Mega Man 10 demo, or even the Sonic CD demo from the PSN store. Press the damn jump button! Is it instant? It sure won't be! Stop testing with 3D games, you'll never know because there are so many factors.

Do you have a laptop?

I played the list in 2D and 3D...no lag

Infamous 2 is not 3D

yes but no HDMI port so cant test the one-foot-snake-oil lets call it OFSO from now on

He meant 3D as in 3D models as in not 2D sprites. He wants you to play a more simple game where the input lag will be more obvious.


Does it have display port or DVI? You can get cheap dongles that will adapt those to HDMI. Do you have friends/family/neighbours who would be so equipped? ENQUIRMING MINDS WANT TO KNOW!

Or, as you're all getting tired of hearing me say, you could get Rock Band 2/3 and test? Shirley you know someone who could loan it to you? Heck, you could buy it used from Game Stop and return within 7 days if you were that against keeping it. Make sure to get an auto-calibrating guitar to go with it. It has a little camera embedded next to the strummer.

Finally, do you have any kind of powered HDMI splitter or switch? This is a stupid test (the whole OFSO thing is stupid, so let's just go with it)... Can you plug your longer cable from the source into the switch/splitter and then the short cable into the 3010 and see what you think? There's no reason the shorter cable should make any difference, but if it does it's really only of any use to me if I can still somehow use the long cable. If repeating the HDMI signal at the switch and then only traversing the OFSO works, then I might have a 3010 in my relative past and near future.

3D games (I don't mean games played with glasses in THREE-DEEEE) are useless for lag tests. They factor in some lag to begin with, and have their own - waiting for vsync, triple-buffering, upscaling, blah blah blah. Using 2D, sprite-based or vector-based games shows you everything you need to know.

It's like those people in the TV forums - "I hate this TV, it shows pixelation when the football players move."

Right, as if the tightly compressed satellite feeds have nothing to do with that! People are mostly ignorant, and see what they see, not knowing what causes it. Using a "television" source to check your display device is asinine, just like using upscaled, 3D, wireless-controller-driven games to test input lag is asinine as well.

Some of you are clearly delusional and are actually hurting people - someone is going to read "Yep, the 3010 has no lag, feels perfect to me, I game all the time and I know what's good and bad," and then buy it, hate it, and can't return it. That's a disservice, and it's not fair to the majority of readers here. Someone is going to drop almost $2000 to find out they've been lied to.

The 3010 and 5010 have horrible lag, end of story. Everyone defending it, please stop. Just because you "can't feel it" doesn't mean it doesn't exist. Do some screenshot proofs and you will see - science beats personal opinion.

I just got an epson 8700 and messed around with uncharted 3 a little. I just stood there and shot bullets. I can definitely see a very slight delay between pushing the button and the bullet coming out, but it's so tiny it will never bother me. That was with Frame Interpolation and Noise Reduction OFF. However, I then turned on Frame Interpolation to "Normal" and did the same thing. BIG difference, the lag was very noticeable and probably would make the game unplayable. Too bad, FI looks pretty awesome on games. I hate it on movies, it looks terrible, but it looks great on games.

Now imagine twice that lag and you've got the new Epsons!