AVS Forum banner
Status
Not open for further replies.
1 - 20 of 32 Posts

·
Registered
Joined
·
61 Posts
Discussion Starter · #1 ·
Hello,


I have just seen the Yamaha DPX-1000 and (unfortunately) it shows motion dithering artifacts like other DLP pjs. So I am beginning to wonder if someday, DLP technology would solve the problem. Any thoughts?

Cheers
 

·
Registered
Joined
·
5,445 Posts
DLP is a fantastic technology that accomplishes a lot with a single chip. Still that is a compromise. With a three chip design color banding and motion dithering should be close to non existent. There is a remedy but you will have to pay for it. Rumor hast started about a Marantz 3-chip hd2.
 

·
Registered
Joined
·
5,445 Posts
Common speculation says presentation at CEDIA and price about 20k$.
 

·
Registered
Joined
·
610 Posts
Quote:
Originally posted by Alan Gouger
Using the DVI input also makes a big difference:)
Alan - Would you explain why the DVI input improves motion dithering?


Thanks
 

·
Registered
Joined
·
61 Posts
Discussion Starter · #9 ·
Jeff,


I have a ISCO II lens and the problem is still very noticeable. Check Matrix beginning of chapter 12 or end of chapter 11. Look at the faces.


Alan,


I use the VGA input via HTPC (I don't think DVI can do much better) and the problem is also very noticeable. Strangely, I noticed that the sharper or the clearer the picture is, the more noticeable is the problem. It is like the mirrors cannot keep up the pace of the picture. Very annoying :(


Cheers
 

·
Registered
Joined
·
1,195 Posts
Quote:
Originally posted by Alan Gouger
Using the DVI input also makes a big difference:)


Yes, that's right, big difference. What also makes for a big difference is reducing the vertical refresh rate. The dithering is caused by too little processing speed of the DMD (at least part of it). As NTSC is (wrongfully so for film material) based on 60Hz there are 12 extra pictures per second to be processed. If you use an HTPC with 48Hz to play NTSC film material the dithering is reduced quite remarkably (besides of course being reliefed of the anoying 3:2 stutters).


Now here is the point: Only the SIM2 HT300+ can do 48Hz via DVI AFAIK (as they have addressed the problem ;) ).


Regards

Christoph



P.S.: Actually 24 fps would be the best thing to do, unfortunately the TI specs do not allow for 24Hz with the HD2 chips.
 

·
Registered
Joined
·
61 Posts
Discussion Starter · #11 ·
Christoph,


As you said, they cannot do 48Hz. But if they can do 72Hz, is it not the same thing (or result)? Same thing for PAL DVDs, with 50Hz or 75Hz.

Cheers


Philippe
 

·
Registered
Joined
·
1,195 Posts
Quote:
Originally posted by Bluematches
But if they can do 72Hz, is it not the same thing (or result)? Same thing for PAL DVDs, with 50Hz or 75Hz.


Philippe,


yes, if they COULD! But simply put, they can't ;) TI specifies the HD2 chip as doing between 48 and 60 Hz.


Lower resolution pjs like the HT200DMF can do 72 and 75Hz, but that is a different story then.


Also the higher the vertical refresh the higher the dithering risk (though at lower resoltion the processing speed seems to be less of an issue).


Cheers

Christoph
 

·
Registered
Joined
·
6,092 Posts
Motion dithering is what many people call temporal dithering and I think that is the official TI name for it as well. There are three things that really help to reduce it.


1. Keep the gamma at reasonable level like 2.2. When you start to use really punchy gammas like 2.5 or higher then the duty cycle of the mirrors gets less and less and dithering becomes more aparent. Motion dithering is most pronounced in dark scenes. At 10 IRE the difference between 2.5 and 2.2 is two fold. With 2.2 the mirrors are on .6% of the time while only .3% for 2.5. The Sim2 HT300+ actually uses a gamma of 2.1 which futher helps out the projector.


2. Brightness is a big factor. The image needs to be in the 25ft-L or less range in intensity to help reduce temporal dithering. The Yamaha is pretty bright. This doesn't mean that bright projectors are bad just that they need to be used on gray screens to get their image intensity correct.


3. Proper analog filtering is very important. DLP projectors will pick up just about anything on the inputs very quickly. If the input filters are not doing their job, then tons of dither looking type garbage is generated by the sampling stage of the projector itself. This is significantly reduced by using DVI.


There are also some other key points like not viewing too close. You should never view an HD-2 from closer than one screen width away and you should normally be 1.5+ screen widths back.


The Sim2 HT300+ does all of these very well and I challange anyone to see dithering in the image it produces. (Its brightness is a much heated debate on this forum, but it is safe to say that it isn't overly bright.)


-Mr. Wigggles
 

·
Registered
Joined
·
3,667 Posts
One of my biggest concerns about DLP is the temporal dithering. If its what I think it is, (general fast moving noise on the image at close range). seems to be very noticeable at short viewing distance and may be responsible for eye fatigue commonly reported.


Just because you can't see it while staring at the screen, doesn't mean that you don't sub-consciously notice it. Like noticing motion out of the corner of your eye.


TI needs to move to 3-chippers on their high end video gear and I'll be looking forward to the Maranz intro. I just hope its bright enough.
 

·
Registered
Joined
·
6,436 Posts
I found that different players can produce different results with regard to the colour banding on faces - out of 4 DVD sources, one did it every time on a particular scene, another did it 2 times out of 10, and two others (inc htpc) didn't do it at all.


All had been calibrated for white and black levels via Avia as a minimum.


The players were Philips 963sa, Toshiba SD520, Limit 9900SE and Radeon 7000 htpc.


Although temporal dithering is part and parcel of DLP, it does seem to be exacerbated by certain players.


Gary.
 

·
Registered
Joined
·
610 Posts
Would somebody explain exactly what the symptom of motion dithering is?


1. Does it make horizontal pans appear to stutter or be jerk rather than be smooth?


2. Does it cause the colors on a panned scene to appear to be banded or otherwise not smooth?


3. Does it cause the dreaded "swarming pixels"?



The only thing I've actually seen on my SP7200 is "1".


Thanks
 

·
Registered
Joined
·
6,092 Posts
dhp,


1. stuttering horizontal pans have nothing to do with temporal dithering. There could be a difficiency in the 7200 but I bet it is the combination of film being only 24 frames per second and your projector being quite bright. My favorite test for stuttering (and temporal dithering) is my Xbox. 480p at 60fps in most situations. If the projector stutters then than it must be a projector problem. Try using your source in a normal CRT TV and look hard and see if you can see the stuttering.


2. Not related to temporal dithering either. Banding is a source issue, a DLP projector that is set-up very poorly or DLP projector that was programmed with horrible gamma tables (I have not seen bad gamma tables in a HD-2 or any other recent DLP)


3. Swarming pixels should only be noticeable when you are sitting very close and are watching a "dark" image. swarming pixels are present all the time because that is how the DMD makes it light output.


Temporal dithering is much more visible "swarming pixels" when motion occurs in the image (once again mainly dark high contrast images). shrink your image down, put on a dark movie scene with characters moving back and forth. Now get close to the image and watch the faces "break-up" when they move from side to side. You should see some level of dithering.


-Mr. Wigggles
 

·
Registered
Joined
·
47 Posts
@clehner


Do you mean that the HD2 units actually have different gradient "resolution" between 48/50 and 60 Hz operation?


I've just been through a full movie with an 7200 for the first time, driving it at 1280x720 @ 60 Hz from a HTPC to the analog VESA in -- and found the temporal dithering and solarisation I've seen quite bad, much worse than I'm used to from my dusty Davis CinemaTen. I seriously hope this is much better at 48/50 Hz, but I didn't manage to drive the unit at such a frequency via the VESA input.


Olli
 

·
Registered
Joined
·
1,195 Posts
Quote:
Originally posted by owagner

[BI've just been through a full movie with an 7200 for the first time, driving it at 1280x720 @ 60 Hz from a HTPC to the analog VESA in -- and found the temporal dithering and solarisation I've seen quite bad, ... [/b]


That is what I have been suspecting with the 7200, thanks for the test.


No, different vertical refreshs have nothing to do with 'gradients of resolution' as you coined it. It has something to do with processing speed. Dithering is partially caused by reducing the color depth at certain points where there isn't enogh time so to speak for processing the whole color palette.


By reducing the number of frames per second your are kind of helping the system. The SIM2 HT300 shows very little dithering at 60Hz, it's practically gone at 48Hz.


Once again I think it's kind of obvious that there is more to a high end DLP than just the HD2 chip ;)


Regards


Christoph
 

·
Registered
Joined
·
61 Posts
Discussion Starter · #20 ·
Mr Wiggles,


So the only DLP pj on the market without temporal dithering is the Sim2 HT300+ ??

I have just seen the Nec HT1000 via DVI. It shows less dithering than any DLP pj I have seen. But it is still there. According to the WSR review of this pj, by changing some options, it can be very close to no dithering. I have to check that.

I am wondering why this problem doesn't seem to be a problem for most of the DLP manufacturers (except Sim apparently). For me, it is the biggest problem on DLP. Is it so complicated to fix? Is a firmware upgrade able to fix the problem on most of the DLPs?

Cheers
 
1 - 20 of 32 Posts
Status
Not open for further replies.
Top