Let's set this straight - No one can do 24p consistently well - Page 21

core i3-2350M
driver 8.15.10.2622

Which search words I must use?
Are custom resolutions useless?

What board do you have? This is a function of the VBIOS.

I have mobile Sandy Bridge. may be here is my problem.
btw, very strange situation. must I check everyone driver now?
Edited by ultrasilent - 7/8/12 at 2:47pm

http://market.yandex.ru/model.xml?modelid=8282216&hid=91013

Best bet is to make sure you have the latest. Other than that, not sure what you can do.

I have latest BIOS and latest video drivers. and I`m not sure that my laptop support downgrading the BIOS.
and what about custom resolution?

Yeah, +- 0.00050 hz isn't so bad - basically closer than previous generations of HTPCs I've built. I could live with that. It sure is weird that the frequencies keep changing on every reboot - I suspect it could be a BIOS problem (in videocard? motherboard?) with respect to PCIe timing. Perhaps having something to do with power management? Oh well. It's more than good enough.

The Dune isn't perfect either - I found one title (Wrath of the Titans) that actually has an audio gap at a seamless branch transition point - but it only occurs on a first playthrough when using DTS-HD MA bitstreaming. If I tell the Dune to decode to PCM first, there is no audio dropout! This would probably not happen on my HTPC (although there are some titles that seem to trip up LAV audio splitter with the same result).

There is no perfect solution. Use what you got, don't die old and bitter!

I can`t understand what problem is with custom resolutions on my Intel HD 3000 video.
I spent some time, try many different presets , but every time system respond with this message:
Your monitor doesn`t support this rate/mode.

where is my mistake?

You're trying to set up a resolution that your monitor doesn't support.

I don`t think so.
same presets on Nvidia work without problem.
are Intel`s custom resolution operational or not?
Can somebody show Intel panel with working custom resolutions?

I'm using custom resolutions with nVidia so IDK as I never used them on the intel driver before I added the nVidia card to my HTPC.

That may be the reason why I can't get those "Intel with UAC off: 23.973 (-0.003)" on my Dell Vostro 3550, i5 (Cougar Point HM67, Sandy Bridge) -> only 23.991 (+0.015)?
1/(23.991-23.976) = each 66.666.. sec I should have a periodic stutter.

To see that I built a new "23.976-Slices" and that's the reason why I'm writing here, would be interesting how it works on your machines.

Download here* 30MB 1080p23.976, H.264, AAC+AC3, 8:22min, low bitrate

Resized 70%:




The "66.666.. sec" didn't really work for me, I have a single stutter every 21 sec, no idea why. Tried it with other clips, decoders, rendrerers, always 21 sec.
Even switching to the Dell built-in HD6630 same 21 sec!? BIOS/driver issue?
Some times an additional micro-stutter at the point the red and green lines are jumping each 66.666.. sec but not always.

Ati HD5850 mobile on the other laptop shows a nice 23.977 and no stutter as far as I could see:





"23.976-Slices" next feature, I built in the "Pong" sounds every time the ticker bouncing (every 5 sec) from the edge.
This beeps are only 50 ms long and so you are able to "see" your video or audio delay rel. exactly (ie. to set the AVR delay):





p.s.
if 8 minutes are too short just set MPC-HC to "Repeat forever" (but not "Rewind..!) and the count down moves on although repeating.

Have a nice time!
.
Edited by blaubart - 8/5/12 at 2:17pm

Hmmm, I used a stop watch to measure the speed of the bouncing Kasescheibchen, and it was about 2.79 seconds to move 10 squares, so rounding that is about 280 msec per square. But I see you have it labeled as 200ms per square.

Doing time calculations programatically is a little more accurate than hitting a button on a stopwatch and rounding the results.

Here is one more data point in our 24p quest. My AMD 6450 video card was taking a nose dive, so I ordered a 7750 (Sapphire, with passive heatsink & no fan). I've only had the 7750 a couple of days now, but so far, it is doing very well. Reclock is reporting exactly 23.976 as I'm playing BDs through TMT5. My 5000 and 6000 series ATI/AMD video cards were slightly off at 23.975 and 23.977, but this 7750 seems to be right on target. Yes, it is overkill for a HTPC, and I wish they'd make cheaper more HTPC oriented video cards, but right now this is the lowest end 7000 series video card they're offering.

i am interested in testing this as i am suspicious that my onkyo avr is messing with the framerate why did you build this as an .au file? it seems mkv would be better for compatibility with more media players (speaking of which are you testing with mpc in your screenshots?)

also with a better container it should be possible to include multiple audio formats i am suspicious that my new avr might have sync issues with dts specifically.

Apparently not, in this case. Your computations appear to be wrong. I just timed it three times and got 2.73, 2.78, 2.82 seconds for 10 squares. There is no way that 200ms per square is correct (at least when played back on my computer with MPC-HC and 23.976Hz display)

By the way, in case you think my reaction time is slow, I closed my eyes and paused playback when I heard the pong sound. I did this several times. Every time the arrow was still within the first square (although close to the edge). So my reaction time, including the time to recognize the sound, press the mouse button, and wait for the video to stop is consistently around 250ms. So even if you assume my 2.8sec per 10 squares time is off by 250ms, it is still no where near 200ms per square (it would be 255ms per square if I was 250ms late).

Well I would check them if they were actually MY computations.

I'm sure blaubart will fix it if is indeed a mistake.

Here is what i am getting from Radeon 6450 how often should i be getting a frame drop?

Slices_1080p23.976_H.264_AAC_audio.jpg 228k .jpg file

You are right, to be precise, one complete ticker movement (to the left + right) is 10.051 sec, so half movement (one direction) is 5.0255 sec

5.0255 / 18 (squares) = 279 ms per square

But what I always try to explain for the patterns is what the eye in real life is usually able to see. And that comes quite close to what I've shown in the picture.
Tested it by changing the audio delay in the Denon AVR to 0, 100 and 200 ms.

-> "pushed" the arrow in the picture 2 pixels to the left, like it more now?





Yes, testing by MPC-HC. No DTS possible with the encoder, sorry. And it HAS to be all done by the encoder cause only he garants true timings.

Tested with various splitter /audio and video decoder /EVR combos - always same results, that's important. mp4 (H.264 + AAC) is a very easy format (nearly?) every player should be able to play.

Looks like a move in the right direction. A quick measurement with my ruler makes it look like you are measuring 214ms delay in the picture labeled 200ms. Close enough.

You hopefully should get no drop. Same with my HD5850 mobile - 23.977 and no drops. Congrats!

May be the test will show drops on Intel Graphics?
- i.e. if somebody had a HD3000 (Cougar Point HM67, Sandy Bridge) like me or later but without HD6630 combined - better then every 21 sec?
- or a newer Ivy Bridge HD4000 where the 24p bug is said to be repaired?

I'm not sure if the formula - 1/(EVR-output - 23.976) = stutters in sec - always gives the right answer. To see if there is more the 23.976-Slices should be handy.


Just try it by changing your AVR's audio delay. Eye measures are quite different.. But may be an ego-shooter pro will see with other eyes ?
Edited by blaubart - 7/29/12 at 2:02pm

I get different results from your slices than any of my movies.

How do you get the display to show how often you will get a dropped frame in MPC?

madVR